Counter Projection - Additional Analyses

Stereotype moderator analysis: Initial Study

#model
projmodel_stereo <- lmer(MFQTarg ~ MFQSelf_c * MFQStereo_c *
                                    inoutgroup_effects * ident_c +
                                    (MFQSelf_c|subid), 
                                 data = stereo_analysis)

#results
tab_model(projmodel_stereo,
          title = "Projection moderated by stereotyping")

Projection moderated by stereotyping

	MFQ Targ
Predictors	Estimates	CI	p
(Intercept)	4.20	4.15 – 4.25	<0.001
MFQSelf_c	0.14	0.12 – 0.17	<0.001
MFQStereo_c	0.07	0.06 – 0.09	<0.001
inoutgroup_effects	0.24	0.13 – 0.34	<0.001
ident_c	0.04	0.01 – 0.07	0.014
MFQSelf_c * MFQStereo_c	-0.03	-0.04 – -0.02	<0.001
MFQSelf_c * inoutgroup_effects	0.37	0.31 – 0.42	<0.001
MFQStereo_c * inoutgroup_effects	0.60	0.57 – 0.63	<0.001
MFQSelf_c * ident_c	0.01	-0.01 – 0.02	0.445
MFQStereo_c * ident_c	-0.03	-0.04 – -0.02	<0.001
inoutgroup_effects * ident_c	0.24	0.18 – 0.30	<0.001
(MFQSelf_c * MFQStereo_c) * inoutgroup_effects	0.01	-0.01 – 0.02	0.405
(MFQSelf_c * MFQStereo_c) * ident_c	-0.01	-0.01 – -0.00	0.005
(MFQSelf_c inoutgroup_effects) ident_c	0.00	-0.03 – 0.03	0.804
(MFQStereo_c inoutgroup_effects) ident_c	0.01	-0.01 – 0.03	0.212
(MFQSelf_c * MFQStereo_c * inoutgroup_effects) * ident_c	-0.00	-0.01 – 0.01	0.956
Random Effects
σ2	1.31
τ00 subid	0.45
τ11 subid.MFQSelf_c	0.08
ρ01 subid	-0.02
ICC	0.33
N subid	674
Observations	20045
Marginal R2 / Conditional R2	0.214 / 0.470

out_only <- stereo_analysis %>% filter(inoutgroup_effects != .5)

projmodel_stereo_out <- lmer(MFQTarg ~ MFQSelf_c * MFQStereo_c * ident_c +
                                    (MFQSelf_c|subid), 
                                 data = out_only)


tab_model(projmodel_stereo_out,
          title = "Projection moderated by stereotyping with out-group data only")

Projection moderated by stereotyping with out-group data only

	MFQ Targ
Predictors	Estimates	CI	p
(Intercept)	4.08	4.00 – 4.16	<0.001
MFQSelf_c	-0.04	-0.08 – 0.01	0.102
MFQStereo_c	-0.23	-0.25 – -0.20	<0.001
ident_c	-0.08	-0.13 – -0.04	<0.001
MFQSelf_c * MFQStereo_c	-0.03	-0.04 – -0.02	<0.001
MFQSelf_c * ident_c	0.00	-0.02 – 0.03	0.759
MFQStereo_c * ident_c	-0.04	-0.05 – -0.02	<0.001
(MFQSelf_c * MFQStereo_c) * ident_c	-0.01	-0.01 – 0.00	0.078
Random Effects
σ2	1.79
τ00 subid	0.46
τ11 subid.MFQSelf_c	0.13
ρ01 subid	0.13
ICC	0.29
N subid	334
Observations	9939
Marginal R2 / Conditional R2	0.052 / 0.331

This analysis was run using only data for the out-group, meaning the main effect of MFQSelf_c now represents the effect of projection for out-group targets only. The effect is not significant, meaning, in the plot below, that the out-group effect is not different from 0.

initial_study_3way <- effect(term = "MFQSelf_c:inoutgroup_effects:ident_c", 
                           xlevels = list(ident_c = c(-1.78, 0, 1.78), 
                                         inoutgroup_effects = c(-.5,.5)), 
                           mod = projmodel_stereo)

initial_study_3way <- as.data.frame(initial_study_3way)

initial_study_3way$Group <- as.factor(initial_study_3way$inoutgroup_effects)
initial_study_3way$Identification <- as.factor(initial_study_3way$ident_c)

ident_label <- c("-1.78" = "Low identification \n (1 SD below)", 
                 "0" = "Average identification",
                 "1.78" = "High identificatiion \n(1 SD above)")

ggplot(initial_study_3way, aes(MFQSelf_c, 
                             fit, 
                             group = Group, 
                             color = Group)) +
  geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black", 
              aes(linetype = Group)) +
  facet_wrap(~Identification, labeller = 
               labeller(Identification = ident_label)) +
  scale_linetype_manual("Group", breaks = c(-.5, .5), 
                       labels = c("Out-group", "In-group"),
                       values = c("solid", 
                                   "dashed")) +
  theme_minimal() +
  labs(title = "Projection predicted by in-group identification and group membership",
       subtitle = "Controlling for stereotyping",
       x = "MFQ responses for self",
       y = "MFQ responses for target")

ggplot(initial_study_main_group, aes(MFQSelf_c, 
                               fit, 
                               group = Group, 
                               color = Group)) +
  geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black",
              aes(linetype = Group)) +
  scale_linetype_manual("Group", breaks = c(-.5, .5), 
                       labels = c("Out-group", "In-group"),
                       values = c("solid", 
                                  "dashed")) +
  theme_minimal() +
  labs(title = "Projection predicted by group membership",
       subtitle = "Controlling for stereotyping",
       x = "MFQ responses for self",
       y = "MFQ responses for target")

These results demonstrate that our significant three-way interaction from before is no longer significant, but the main effect of group identification still is significant and still shows counter projection. The three-way is not significant because, regardless of in-group identification, people are still counter projecting to the same extent with their out-group and projecting normally with their in-group.

With political affiliation as moderator

stereo_analysis %>% 
  select(subid, par_cand) %>% 
  unique() %>% 
  group_by(par_cand) %>% 
  count() %>% 
  ungroup() %>% 
  mutate(total = sum(n),
         percent = (n/total)*100)

## # A tibble: 2 x 4
##   par_cand     n total percent
##   <fct>    <int> <int>   <dbl>
## 1 Clinton    429   674    63.6
## 2 Trump      245   674    36.4

is_cand <- lmer(MFQTarg ~ MFQSelf_c * MFQStereo_c *
                                    inoutgroup_effects * ident_c * par_cand +
                                    (MFQSelf_c|subid), 
                                 data = stereo_analysis)

tab_model(is_cand,
          title = "Projection moderated by stereotyping")

Projection moderated by stereotyping

	MFQ Targ
Predictors	Estimates	CI	p
(Intercept)	4.30	4.22 – 4.37	<0.001
MFQSelf_c	0.10	0.06 – 0.14	<0.001
MFQStereo_c	0.12	0.09 – 0.14	<0.001
inoutgroup_effects	0.01	-0.14 – 0.17	0.847
ident_c	0.10	0.06 – 0.15	<0.001
par_cand [Trump]	-0.17	-0.30 – -0.04	0.009
MFQSelf_c * MFQStereo_c	-0.04	-0.05 – -0.02	<0.001
MFQSelf_c * inoutgroup_effects	0.51	0.44 – 0.59	<0.001
MFQStereo_c * inoutgroup_effects	0.62	0.58 – 0.67	<0.001
MFQSelf_c * ident_c	-0.00	-0.02 – 0.02	0.915
MFQStereo_c * ident_c	-0.02	-0.03 – -0.01	0.002
inoutgroup_effects * ident_c	0.15	0.05 – 0.25	0.003
MFQSelf_c * par_cand [Trump]	0.13	0.07 – 0.19	<0.001
MFQStereo_c * par_cand [Trump]	-0.10	-0.13 – -0.06	<0.001
inoutgroup_effects * par_cand [Trump]	0.53	0.27 – 0.79	<0.001
ident_c * par_cand [Trump]	-0.08	-0.14 – -0.01	0.033
(MFQSelf_c * MFQStereo_c) * inoutgroup_effects	0.06	0.04 – 0.08	<0.001
(MFQSelf_c * MFQStereo_c) * ident_c	-0.02	-0.03 – -0.01	<0.001
(MFQSelf_c inoutgroup_effects) ident_c	0.07	0.02 – 0.11	0.003
(MFQStereo_c inoutgroup_effects) ident_c	0.01	-0.02 – 0.03	0.567
(MFQSelf_c * MFQStereo_c) * par_cand [Trump]	-0.01	-0.03 – 0.01	0.474
(MFQSelf_c inoutgroup_effects) par_cand [Trump]	-0.38	-0.50 – -0.26	<0.001
(MFQStereo_c inoutgroup_effects) par_cand [Trump]	-0.10	-0.17 – -0.03	0.007
(MFQSelf_c * ident_c) * par_cand [Trump]	-0.01	-0.05 – 0.02	0.364
(MFQStereo_c * ident_c) * par_cand [Trump]	-0.00	-0.02 – 0.02	0.822
(inoutgroup_effects ident_c) par_cand [Trump]	0.03	-0.11 – 0.17	0.651
(MFQSelf_c * MFQStereo_c * inoutgroup_effects) * ident_c	0.03	0.02 – 0.04	<0.001
(MFQSelf_c * MFQStereo_c * inoutgroup_effects) * par_cand [Trump]	-0.11	-0.14 – -0.07	<0.001
(MFQSelf_c * MFQStereo_c * ident_c) * par_cand [Trump]	0.02	0.01 – 0.03	<0.001
(MFQSelf_c inoutgroup_effects ident_c) * par_cand [Trump]	-0.03	-0.09 – 0.03	0.370
(MFQStereo_c inoutgroup_effects ident_c) * par_cand [Trump]	0.02	-0.02 – 0.05	0.353
(MFQSelf_c * MFQStereo_c * inoutgroup_effects ident_c) par_cand [Trump]	-0.03	-0.05 – -0.01	0.001
Random Effects
σ2	1.30
τ00 subid	0.44
τ11 subid.MFQSelf_c	0.08
ρ01 subid	0.03
ICC	0.32
N subid	674
Observations	20045
Marginal R2 / Conditional R2	0.234 / 0.477

When we include who the participant voted for with stereotyping as well, we still have the main effect from above. Now, the interaction between target group and in-group identification is significant. There is also an interaction between target group (in or out) and who the participant voted for. However, an interaction between target group, degree of in-group identification, and who they voted for is not significant. There are other significant interactions including stereotyping, but, since we only included it as a moderator to see if the relationship persisted, I didn’t go through them thoroughly.

is_group_cand <- effect(term = "MFQSelf_c:inoutgroup_effects:par_cand", 
                             xlevels = list(inoutgroup_effects = c(-.5,.5),
                                            par_cand = c("Clinton", "Trump")), 
                             mod = is_cand)

is_group_cand <- as.data.frame(is_group_cand)

is_group_cand$Group <- as.factor(is_group_cand$inoutgroup_effects)

ggplot(is_group_cand, aes(MFQSelf_c, 
                               fit, 
                               group = Group, 
                               color = Group)) +
  geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black",
              aes(linetype = Group)) +
  facet_wrap(~par_cand)+
  scale_linetype_manual("Group", breaks = c(-.5, .5), 
                       labels = c("Out-group", "In-group"),
                       values = c("solid", 
                                  "dashed")) +
  theme_minimal() +
  labs(title = "Projection predicted by group membership",
       subtitle = "Controlling for stereotyping",
       x = "MFQ responses for self",
       y = "MFQ responses for target")

Clinton supporters counter project and Trump supporters do not. I’ve checked this a lot and it seems to be correct, I must have messed up in my first year when I analyzed it…

is_3way_cand <- effect(term = "MFQSelf_c:inoutgroup_effects:ident_c", 
                           xlevels = list(ident_c = c(-1.78, 0, 1.78), 
                                         inoutgroup_effects = c(-.5,.5)), 
                           mod = is_cand)

is_3way_cand <- as.data.frame(is_3way_cand)

is_3way_cand$Group <- as.factor(is_3way_cand$inoutgroup_effects)
is_3way_cand$Identification <- as.factor(is_3way_cand$ident_c)

ident_label <- c("-1.78" = "Low identification \n (1 SD below)", 
                 "0" = "Average identification",
                 "1.78" = "High identificatiion \n(1 SD above)")

ggplot(is_3way_cand, aes(MFQSelf_c, 
                             fit, 
                             group = Group, 
                             color = Group)) +
  geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black", 
              aes(linetype = Group)) +
  facet_wrap(~Identification, labeller = 
               labeller(Identification = ident_label)) +
  scale_linetype_manual("Group", breaks = c(-.5, .5), 
                       labels = c("Out-group", "In-group"),
                       values = c("solid", 
                                   "dashed")) +
  theme_minimal() +
  labs(title = "Projection predicted by in-group identification and group membership",
       subtitle = "Controlling for stereotyping",
       x = "MFQ responses for self",
       y = "MFQ responses for target")

ggplot(is_group_cand, aes(MFQSelf_c, 
                               fit, 
                               group = Group, 
                               color = Group)) +
  geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black",
              aes(linetype = Group)) +
  scale_linetype_manual("Group", breaks = c(-.5, .5), 
                       labels = c("Out-group", "In-group"),
                       values = c("solid", 
                                  "dashed")) +
  theme_minimal() +
  labs(title = "Projection predicted by group membership",
       subtitle = "Controlling for stereotyping",
       x = "MFQ responses for self",
       y = "MFQ responses for target")

MFQ sub areas

#separating MFQ into binding and individualizing subscales
mfq_binding <- clean_data2 %>% 
  filter(MFQ_Number %in% c("3", "4", "5", "9", "10", "11", "14", "15", "16", 
                           "19", "20", "21", "25", "26", "27", "30", "31", "32")) %>%
  select(subid, MFQ_Number, MFQTarg, MFQSelf_c, MFQSelf, MFQStereo_c, MFQStereo, inoutgroup_effects, ident_c, par_cand, Liking, WilltoDiscuss, Valid, Treatment, Contentious) %>% 
  unique() %>% 
  na.omit()

mfq_individual <- clean_data2 %>% 
  filter(MFQ_Number %in% c("1", "2", "7", "8", "12", "13", "17", "18", "23", 
                           "24", "28", "29")) %>%
  select(subid, MFQ_Number, MFQTarg, MFQSelf_c, MFQSelf, MFQStereo_c, MFQStereo, inoutgroup_effects, ident_c, par_cand, Liking, WilltoDiscuss, Valid, Treatment, Contentious) %>% 
  unique() %>% 
  na.omit()

Initial Study

Binding

#model
is_mfqbinding <- lmer(MFQTarg ~ MFQSelf_c * MFQStereo_c *
                                    inoutgroup_effects * ident_c +
                                    (MFQSelf_c|subid), 
                                 data = mfq_binding)

#results
tab_model(is_mfqbinding,
          title = "Projection moderated by stereotyping for binding morals")

Projection moderated by stereotyping for binding morals

	MFQ Targ
Predictors	Estimates	CI	p
(Intercept)	4.19	4.12 – 4.27	<0.001
MFQSelf_c	0.13	0.10 – 0.15	<0.001
MFQStereo_c	0.15	0.13 – 0.17	<0.001
inoutgroup_effects	-0.06	-0.21 – 0.09	0.416
ident_c	-0.02	-0.06 – 0.02	0.424
MFQSelf_c * MFQStereo_c	-0.02	-0.04 – -0.01	<0.001
MFQSelf_c * inoutgroup_effects	0.30	0.25 – 0.35	<0.001
MFQStereo_c * inoutgroup_effects	0.30	0.26 – 0.34	<0.001
MFQSelf_c * ident_c	0.00	-0.01 – 0.01	0.807
MFQStereo_c * ident_c	-0.03	-0.04 – -0.02	<0.001
inoutgroup_effects * ident_c	0.43	0.35 – 0.51	<0.001
(MFQSelf_c * MFQStereo_c) * inoutgroup_effects	-0.04	-0.06 – -0.02	<0.001
(MFQSelf_c * MFQStereo_c) * ident_c	-0.01	-0.01 – -0.00	0.006
(MFQSelf_c inoutgroup_effects) ident_c	0.04	0.01 – 0.06	0.007
(MFQStereo_c inoutgroup_effects) ident_c	0.03	0.01 – 0.05	0.007
(MFQSelf_c * MFQStereo_c * inoutgroup_effects) * ident_c	0.01	-0.01 – 0.02	0.364
Random Effects
σ2	1.00
τ00 subid	0.90
τ11 subid.MFQSelf_c	0.05
ρ01 subid	0.04
ICC	0.50
N subid	669
Observations	11942
Marginal R2 / Conditional R2	0.189 / 0.593

is_binding_3way <- effect(term = "MFQSelf_c:inoutgroup_effects:ident_c", 
                           xlevels = list(ident_c = c(-1.78, 0, 1.78), 
                                         inoutgroup_effects = c(-.5,.5)), 
                           mod = is_mfqbinding)

is_binding_3way <- as.data.frame(is_binding_3way)

is_binding_3way$Group <- as.factor(is_binding_3way$inoutgroup_effects)
is_binding_3way$Identification <- as.factor(is_binding_3way$ident_c)

ident_label <- c("-1.78" = "Low identification \n (1 SD below)", 
                 "0" = "Average identification",
                 "1.78" = "High identificatiion \n(1 SD above)")

ggplot(is_binding_3way, aes(MFQSelf_c,
                             fit, 
                             group = Group,
                             color = Group)) +
  geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black", 
              aes(linetype = Group)) +
  facet_wrap(~Identification, labeller = 
               labeller(Identification = ident_label)) +
  scale_linetype_manual("Group", breaks = c(-.5, .5), 
                       labels = c("Out-group", "In-group"),
                       values = c("solid", 
                                   "dashed")) +
  theme_minimal() +
  labs(title = "Projection of binding morals predicted by in-group identification and group membership",
       subtitle = "Controlling for stereotyping",
       x = "MFQ responses for self",
       y = "MFQ responses for target")

ggplot(is_binding_group, aes(MFQSelf_c, 
                               fit, 
                               group = Group, 
                               color = Group)) +
  geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black",
              aes(linetype = Group)) +
  scale_linetype_manual("Group", breaks = c(-.5, .5), 
                       labels = c("Out-group", "In-group"),
                       values = c("solid", 
                                  "dashed")) +
  theme_minimal() +
  labs(title = "Projection of binding morals predicted by group membership",
       subtitle = "Controlling for stereotyping",
       x = "MFQ responses for self",
       y = "MFQ responses for target")

Binding including participant vote

is_mfqbinding_cand <- lmer(MFQTarg ~ MFQSelf_c * MFQStereo_c *
                                    inoutgroup_effects * ident_c * par_cand +
                                    (MFQSelf_c|subid), 
                                 data = mfq_binding)

tab_model(is_mfqbinding_cand,
          title = "Projection moderated by stereotyping for binding morals")

Projection moderated by stereotyping for binding morals

	MFQ Targ
Predictors	Estimates	CI	p
(Intercept)	4.46	4.37 – 4.56	<0.001
MFQSelf_c	0.12	0.09 – 0.15	<0.001
MFQStereo_c	0.20	0.17 – 0.23	<0.001
inoutgroup_effects	-0.71	-0.90 – -0.53	<0.001
ident_c	0.12	0.06 – 0.18	<0.001
par_cand [Trump]	-0.59	-0.75 – -0.43	<0.001
MFQSelf_c * MFQStereo_c	-0.02	-0.03 – -0.00	0.026
MFQSelf_c * inoutgroup_effects	0.33	0.26 – 0.40	<0.001
MFQStereo_c * inoutgroup_effects	0.30	0.24 – 0.35	<0.001
MFQSelf_c * ident_c	0.01	-0.01 – 0.03	0.530
MFQStereo_c * ident_c	-0.02	-0.04 – -0.00	0.033
inoutgroup_effects * ident_c	0.19	0.07 – 0.31	0.002
MFQSelf_c * par_cand [Trump]	0.06	-0.00 – 0.11	0.052
MFQStereo_c * par_cand [Trump]	-0.15	-0.19 – -0.10	<0.001
inoutgroup_effects * par_cand [Trump]	1.79	1.46 – 2.11	<0.001
ident_c * par_cand [Trump]	-0.10	-0.18 – -0.01	0.028
(MFQSelf_c * MFQStereo_c) * inoutgroup_effects	-0.03	-0.07 – -0.00	0.028
(MFQSelf_c * MFQStereo_c) * ident_c	-0.01	-0.02 – -0.00	0.003
(MFQSelf_c inoutgroup_effects) ident_c	0.06	0.02 – 0.10	0.008
(MFQStereo_c inoutgroup_effects) ident_c	0.03	-0.00 – 0.07	0.059
(MFQSelf_c * MFQStereo_c) * par_cand [Trump]	-0.04	-0.06 – -0.01	0.005
(MFQSelf_c inoutgroup_effects) par_cand [Trump]	-0.05	-0.16 – 0.06	0.370
(MFQStereo_c inoutgroup_effects) par_cand [Trump]	0.03	-0.06 – 0.12	0.531
(MFQSelf_c * ident_c) * par_cand [Trump]	-0.03	-0.06 – 0.00	0.098
(MFQStereo_c * ident_c) * par_cand [Trump]	0.01	-0.01 – 0.04	0.255
(inoutgroup_effects ident_c) par_cand [Trump]	-0.01	-0.19 – 0.16	0.881
(MFQSelf_c * MFQStereo_c * inoutgroup_effects) * ident_c	0.01	-0.01 – 0.03	0.283
(MFQSelf_c * MFQStereo_c * inoutgroup_effects) * par_cand [Trump]	0.00	-0.05 – 0.06	0.883
(MFQSelf_c * MFQStereo_c * ident_c) * par_cand [Trump]	0.02	0.01 – 0.03	0.003
(MFQSelf_c inoutgroup_effects ident_c) * par_cand [Trump]	-0.03	-0.09 – 0.03	0.369
(MFQStereo_c inoutgroup_effects ident_c) * par_cand [Trump]	-0.02	-0.07 – 0.03	0.416
(MFQSelf_c * MFQStereo_c * inoutgroup_effects ident_c) par_cand [Trump]	-0.01	-0.04 – 0.02	0.524
Random Effects
σ2	0.99
τ00 subid	0.70
τ11 subid.MFQSelf_c	0.04
ρ01 subid	0.06
ICC	0.44
N subid	669
Observations	11942
Marginal R2 / Conditional R2	0.284 / 0.599

Main effect of target group is still significant. Who participants voted for is marginal (.050). The interaction between target group and identification is still significant. The interaction between target group, identification, and who they voted for is not significant.

bind_group_cand <- effect(term = "MFQSelf_c:inoutgroup_effects:par_cand", 
                             xlevels = list(inoutgroup_effects = c(-.5,.5),
                                            par_cand = c("Clinton", "Trump")), 
                             mod = is_mfqbinding_cand)

bind_group_cand <- as.data.frame(bind_group_cand)

bind_group_cand$Group <- as.factor(bind_group_cand$inoutgroup_effects)

ggplot(bind_group_cand, aes(MFQSelf_c, 
                               fit, 
                               group = Group, 
                               color = Group)) +
  geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black",
              aes(linetype = Group)) +
  facet_wrap(~par_cand)+
  scale_linetype_manual("Group", breaks = c(-.5, .5), 
                       labels = c("Out-group", "In-group"),
                       values = c("solid", 
                                  "dashed")) +
  theme_minimal() +
  labs(title = "Projection predicted by group membership",
       subtitle = "Controlling for stereotyping",
       x = "MFQ responses for self",
       y = "MFQ responses for target")

Individualizing

is_mfqindividual <- lmer(MFQTarg ~ MFQSelf_c * MFQStereo_c *
                                    inoutgroup_effects * ident_c +
                                    (MFQSelf_c|subid), 
                                 data = mfq_individual)

tab_model(is_mfqindividual,
          title = "Projection moderated by stereotyping for individualizing morals")

Projection moderated by stereotyping for individualizing morals

	MFQ Targ
Predictors	Estimates	CI	p
(Intercept)	4.16	4.08 – 4.23	<0.001
MFQSelf_c	0.21	0.18 – 0.24	<0.001
MFQStereo_c	0.15	0.13 – 0.17	<0.001
inoutgroup_effects	0.85	0.70 – 1.00	<0.001
ident_c	0.13	0.09 – 0.17	<0.001
MFQSelf_c * MFQStereo_c	-0.03	-0.04 – -0.02	<0.001
MFQSelf_c * inoutgroup_effects	0.24	0.18 – 0.30	<0.001
MFQStereo_c * inoutgroup_effects	0.30	0.25 – 0.35	<0.001
MFQSelf_c * ident_c	-0.01	-0.03 – 0.00	0.093
MFQStereo_c * ident_c	-0.02	-0.03 – -0.01	0.003
inoutgroup_effects * ident_c	-0.06	-0.14 – 0.03	0.186
(MFQSelf_c * MFQStereo_c) * inoutgroup_effects	-0.06	-0.08 – -0.04	<0.001
(MFQSelf_c * MFQStereo_c) * ident_c	0.00	-0.00 – 0.01	0.711
(MFQSelf_c inoutgroup_effects) ident_c	0.02	-0.01 – 0.05	0.170
(MFQStereo_c inoutgroup_effects) ident_c	0.03	0.01 – 0.06	0.009
(MFQSelf_c * MFQStereo_c * inoutgroup_effects) * ident_c	0.01	-0.01 – 0.02	0.305
Random Effects
σ2	0.87
τ00 subid	0.86
τ11 subid.MFQSelf_c	0.06
ρ01 subid	0.01
ICC	0.54
N subid	669
Observations	7956
Marginal R2 / Conditional R2	0.240 / 0.649

The main effect of target group is still significant. The interaction between target group and identification is no longer significant, but I plotted it just to show the difference.

describe(mfq_individual$ident_c)

##    vars    n  mean   sd median trimmed mad   min  max range  skew kurtosis   se
## X1    1 7956 -0.03 1.78  -0.01   -0.04 1.8 -4.23 3.77     8 -0.02    -0.32 0.02

is_indiv_3way <- effect(term = "MFQSelf_c:inoutgroup_effects:ident_c", 
                           xlevels = list(ident_c = c(-1.78, 0, 1.78), 
                                         inoutgroup_effects = c(-.5,.5)), 
                           mod = is_mfqindividual)

is_indiv_3way <- as.data.frame(is_indiv_3way)

is_indiv_3way$Group <- as.factor(is_indiv_3way$inoutgroup_effects)
is_indiv_3way$Identification <- as.factor(is_indiv_3way$ident_c)

ident_label <- c("-1.78" = "Low identification \n (1 SD below)", 
                 "0" = "Average identification",
                 "1.78" = "High identificatiion \n(1 SD above)")

ggplot(is_indiv_3way, aes(MFQSelf_c, 
                             fit, 
                             group = Group, 
                             color = Group)) +
  geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black", 
              aes(linetype = Group)) +
  facet_wrap(~Identification, labeller = 
               labeller(Identification = ident_label)) +
  scale_linetype_manual("Group", breaks = c(-.5, .5), 
                       labels = c("Out-group", "In-group"),
                       values = c("solid", 
                                   "dashed")) +
  theme_minimal() +
  labs(title = "Projection of individualizing morals predicted by in-group identification and group membership",
       subtitle = "Controlling for stereotyping",
       x = "MFQ responses for self",
       y = "MFQ responses for target")

ggplot(is_indiv_group, aes(MFQSelf_c, 
                               fit, 
                               group = Group, 
                               color = Group)) +
  geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black",
              aes(linetype = Group)) +
  scale_linetype_manual("Group", breaks = c(-.5, .5), 
                       labels = c("Out-group", "In-group"),
                       values = c("solid", 
                                  "dashed")) +
  theme_minimal() +
  labs(title = "Projection of individualizing morals predicted by group membership",
       subtitle = "Controlling for stereotyping",
       x = "MFQ responses for self",
       y = "MFQ responses for target")

Individualizing including participant vote

#model
is_mfqindividual_cand <- lmer(MFQTarg ~ MFQSelf_c * MFQStereo_c *
                                    inoutgroup_effects * ident_c * par_cand +
                                    (MFQSelf_c|subid), 
                                 data = mfq_individual)

#results
tab_model(is_mfqindividual_cand,
          title = "Projection moderated by stereotyping for individual morals")

Projection moderated by stereotyping for individual morals

	MFQ Targ
Predictors	Estimates	CI	p
(Intercept)	3.92	3.82 – 4.02	<0.001
MFQSelf_c	0.23	0.19 – 0.27	<0.001
MFQStereo_c	0.19	0.15 – 0.22	<0.001
inoutgroup_effects	1.56	1.36 – 1.76	<0.001
ident_c	0.04	-0.02 – 0.11	0.224
par_cand [Trump]	0.64	0.47 – 0.81	<0.001
MFQSelf_c * MFQStereo_c	-0.03	-0.05 – -0.01	0.001
MFQSelf_c * inoutgroup_effects	0.28	0.19 – 0.36	<0.001
MFQStereo_c * inoutgroup_effects	0.15	0.07 – 0.22	<0.001
MFQSelf_c * ident_c	-0.01	-0.03 – 0.02	0.582
MFQStereo_c * ident_c	0.00	-0.02 – 0.03	0.756
inoutgroup_effects * ident_c	0.21	0.08 – 0.34	0.001
MFQSelf_c * par_cand [Trump]	-0.06	-0.13 – 0.01	0.077
MFQStereo_c * par_cand [Trump]	-0.03	-0.09 – 0.02	0.244
inoutgroup_effects * par_cand [Trump]	-1.80	-2.14 – -1.46	<0.001
ident_c * par_cand [Trump]	-0.01	-0.10 – 0.08	0.775
(MFQSelf_c * MFQStereo_c) * inoutgroup_effects	-0.02	-0.06 – 0.02	0.272
(MFQSelf_c * MFQStereo_c) * ident_c	-0.00	-0.02 – 0.01	0.457
(MFQSelf_c inoutgroup_effects) ident_c	0.04	-0.02 – 0.09	0.167
(MFQStereo_c inoutgroup_effects) ident_c	-0.03	-0.08 – 0.01	0.130
(MFQSelf_c * MFQStereo_c) * par_cand [Trump]	-0.02	-0.05 – 0.01	0.184
(MFQSelf_c inoutgroup_effects) par_cand [Trump]	-0.06	-0.21 – 0.08	0.370
(MFQStereo_c inoutgroup_effects) par_cand [Trump]	0.19	0.09 – 0.30	<0.001
(MFQSelf_c * ident_c) * par_cand [Trump]	0.01	-0.03 – 0.05	0.596
(MFQStereo_c * ident_c) * par_cand [Trump]	-0.03	-0.06 – -0.00	0.042
(inoutgroup_effects ident_c) par_cand [Trump]	-0.03	-0.21 – 0.15	0.746
(MFQSelf_c * MFQStereo_c * inoutgroup_effects) * ident_c	0.03	0.01 – 0.06	0.006
(MFQSelf_c * MFQStereo_c * inoutgroup_effects) * par_cand [Trump]	-0.02	-0.08 – 0.03	0.443
(MFQSelf_c * MFQStereo_c * ident_c) * par_cand [Trump]	0.01	-0.00 – 0.03	0.157
(MFQSelf_c inoutgroup_effects ident_c) * par_cand [Trump]	-0.04	-0.11 – 0.04	0.331
(MFQStereo_c inoutgroup_effects ident_c) * par_cand [Trump]	0.07	0.01 – 0.13	0.016
(MFQSelf_c * MFQStereo_c * inoutgroup_effects ident_c) par_cand [Trump]	-0.04	-0.07 – -0.01	0.010
Random Effects
σ2	0.87
τ00 subid	0.69
τ11 subid.MFQSelf_c	0.06
ρ01 subid	0.03
ICC	0.49
N subid	669
Observations	7956
Marginal R2 / Conditional R2	0.320 / 0.654

Main effect of target group is still significant. The main effect of who participants voted for is marginal (.077). The interactions between target group and identification between target group and who participants voted for, and target group, identification, and who participants voted for are all not significant.

ggplot(is_indiv_group_cand, aes(MFQSelf_c, 
                               fit, 
                               group = Group, 
                               color = Group)) +
  geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black",
              aes(linetype = Group)) +
  scale_linetype_manual("Group", breaks = c(-.5, .5), 
                       labels = c("Out-group", "In-group"),
                       values = c("solid", 
                                  "dashed")) +
  theme_minimal() +
  labs(title = "Projection of individualizing morals predicted by group membership",
       subtitle = "Controlling for stereotyping",
       x = "MFQ responses for self",
       y = "MFQ responses for target")

Follow-up 1

#separating MFQ into binding and individualizing subscales
f1_mfq_binding <- pptnpf1_comb %>% 
  filter(number %in% c("3", "4", "5", "9", "10", "11", "14", "15", "16", 
                           "20", "21", "22")) %>%
  select(Subject, number, Condition, MFQSelf_c, MFQSelf, Identification_c, Gender, Race, MFQTarg) %>% 
  unique() %>% 
  na.omit()

f1_mfq_individual <- pptnpf1_comb %>% 
  filter(number %in% c("1", "2", "7", "8", "12", "13", "18", "19")) %>%
  select(Subject, number, Condition, MFQSelf_c, MFQSelf, Identification_c, Gender, Race, MFQTarg) %>% 
  unique() %>% 
  na.omit()

Binding

f1_binding <- lmer(MFQTarg ~ MFQSelf_c*Identification_c*Condition + 
                              (MFQSelf_c|Subject), 
                            data = f1_mfq_binding, na.action = na.omit)

tab_model(f1_binding,
          title = "Effects of predictors and interactions on projection of binding morals")

Effects of predictors and interactions on projection of binding morals

	MFQ Targ
Predictors	Estimates	CI	p
(Intercept)	4.35	4.25 – 4.45	<0.001
MFQSelf_c	0.12	0.06 – 0.17	<0.001
Identification_c	0.19	0.10 – 0.28	<0.001
ConditionPRWvPB	-0.02	-0.09 – 0.05	0.596
ConditionPRvW	0.07	-0.01 – 0.15	0.097
ConditionPvR	-0.07	-0.22 – 0.08	0.392
MFQSelf_c * Identification_c	-0.08	-0.13 – -0.03	0.001
MFQSelf_c : ConditionPRWvPB	0.07	0.04 – 0.11	<0.001
MFQSelf_c : ConditionPRvW	0.13	0.09 – 0.17	<0.001
MFQSelf_c : ConditionPvR	0.15	0.07 – 0.23	<0.001
Identification_c : ConditionPRWvPB	-0.06	-0.11 – -0.01	0.027
Identification_c : ConditionPRvW	-0.05	-0.12 – 0.02	0.196
Identification_c : ConditionPvR	0.25	0.11 – 0.39	<0.001
MFQSelf_c : Identification_c : ConditionPRWvPB	0.03	0.00 – 0.05	0.024
MFQSelf_c : Identification_c : ConditionPRvW	0.02	-0.02 – 0.06	0.257
MFQSelf_c : Identification_c : ConditionPvR	0.07	0.00 – 0.14	0.047
Random Effects
σ2	1.24
τ00 Subject	0.50
τ11 Subject.MFQSelf_c	0.09
ρ01 Subject	-0.13
ICC	0.37
N Subject	306
Observations	3057
Marginal R2 / Conditional R2	0.100 / 0.436

#running the plot

ggpreview <- function(...) {
  fname <- tempfile(fileext = ".png")
  ggsave(filename = fname, ...)
  system2("open", fname)
  invisible(NULL)
}

knitr::opts_chunk$set(fig.width = 8, fig.height = 6) 

ggplot(f1_binding_summary_3way, aes(x = MFQSelf_c, 
                         y = lsmean, 
                         group = as.factor(Condition))) + 
  facet_grid(.~Identification_c,
             labeller = labeller(Identification_c = Ident_label)) + 
  geom_line(size = .7, aes(linetype = Condition)) +
  xlab("MFQ responses for self") + 
  ylab("MFQ responses for target") + 
  scale_linetype_manual("Issue polarization", 
                        breaks = c(1,2,3,4), 
                        labels = c ("Political\ncandidate",
                                   "Racial\ndisadvantage",
                                   "Marijuana\nlegalization",
                                   "Peanut\nbutter"),
                        values = c("solid",
                                   "twodash", 
                                   "dotted", 
                                   "dashed")) + 
  ggtitle("Projection predicted by identification and issue polarization") +
  theme_minimal(base_size = 13) +
  theme(legend.text = element_text(size = 10),
        legend.key.width = unit(1, "cm"),
        legend.position = "bottom")

ggpreview(width = 8, height = 6, units = "in")

Not much different than the overall, except now average identifiers in the political condition are negatively projecting and people in the racial disadvantage condition who highly identify aren’t projecting at all.

knitr::opts_chunk$set(fig.width = 8, fig.height = 6) 

ggplot(f1_binding_sum_cond, aes(x = MFQSelf_c, 
                              y = lsmean, 
                              group = as.factor(Condition)))+
  geom_line(size = 1, aes(linetype = Condition)) +
  xlab("MFQ responses for self") + 
  ylab("MFQ responses for target") + 
  scale_linetype_manual("Issue polarization", 
                      breaks = c(1,2,3,4),
                      labels = c("Political\ncandidate",
                                  "Racial\ndisadvantage", 
                                  "Marijuana\nlegalization",
                                  "Peanut\nbutter"),
                       values = c("solid",
                                   "twodash", 
                                   "dotted", 
                                   "dashed")) +
  ggtitle("Effect of issue polarization on projection with out-group targets") +
  theme_minimal(base_size = 13) +
  theme(legend.text = element_text(size = 10),
        legend.key.width = unit(1.2, "cm"))

ggpreview(width = 8, height = 6, units = "in")

Only difference is now there is counter projection in the political candidate condition.

knitr::opts_chunk$set(fig.width = 8, fig.height = 6) 

ggplot(f1_binding_ident, aes(MFQSelf_c, 
                               fit, 
                               group = Identification)) +
  geom_line(aes(linetype = Identification), size = .7) +
  scale_linetype_manual("Identification", 
                       breaks = c(-1.19, 0, 1.19), 
                       labels = c ("Low (1 SD below)", 
                                  "Average ", 
                                  "High (1 SD above)"),
                        values = c("solid",
                                   "twodash", 
                                   "dotted")) +
  theme_minimal(base_size = 13) +
  labs(title = "Projection predicted by degree of in-group identification",
       x = "MFQ responses for self",
       y = "MFQ responses for target")

ggpreview(width = 8, height = 6, units = "in")

Pretty similar to the overall model.

Individualizing

f1_individual <- lmer(MFQTarg ~ MFQSelf_c*Identification_c*Condition + 
                              (MFQSelf_c|Subject), 
                            data = f1_mfq_individual, na.action = na.omit)

tab_model(f1_individual,
          title = "Effects of predictors and interactions on projection of individualizing morals")

Effects of predictors and interactions on projection of individualizing morals

	MFQ Targ
Predictors	Estimates	CI	p
(Intercept)	4.13	4.03 – 4.22	<0.001
MFQSelf_c	0.07	0.01 – 0.13	0.033
Identification_c	0.07	-0.01 – 0.16	0.095
ConditionPRWvPB	0.03	-0.03 – 0.09	0.316
ConditionPRvW	0.13	0.06 – 0.20	0.001
ConditionPvR	-0.01	-0.15 – 0.13	0.850
MFQSelf_c * Identification_c	-0.15	-0.21 – -0.09	<0.001
MFQSelf_c : ConditionPRWvPB	0.10	0.05 – 0.14	<0.001
MFQSelf_c : ConditionPRvW	0.17	0.12 – 0.22	<0.001
MFQSelf_c : ConditionPvR	0.22	0.13 – 0.32	<0.001
Identification_c : ConditionPRWvPB	-0.01	-0.06 – 0.04	0.626
Identification_c : ConditionPRvW	0.01	-0.06 – 0.07	0.870
Identification_c : ConditionPvR	0.25	0.12 – 0.38	<0.001
MFQSelf_c : Identification_c : ConditionPRWvPB	0.04	0.01 – 0.07	0.014
MFQSelf_c : Identification_c : ConditionPRvW	0.09	0.04 – 0.13	<0.001
MFQSelf_c : Identification_c : ConditionPvR	0.02	-0.07 – 0.11	0.681
Random Effects
σ2	1.29
τ00 Subject	0.37
τ11 Subject.MFQSelf_c	0.17
ρ01 Subject	0.06
ICC	0.38
N Subject	306
Observations	2446
Marginal R2 / Conditional R2	0.132 / 0.460

The results of this model are also not that different from the overall model: the main effects of condition and identification are still significant, as well as the interaction between condition and identification.

#running the plot

ggpreview <- function(...) {
  fname <- tempfile(fileext = ".png")
  ggsave(filename = fname, ...)
  system2("open", fname)
  invisible(NULL)
}

knitr::opts_chunk$set(fig.width = 8, fig.height = 6) 

ggplot(f1_individual_summary_3way, aes(x = MFQSelf_c, 
                         y = lsmean, 
                         group = as.factor(Condition))) + 
  facet_grid(.~Identification_c,
             labeller = labeller(Identification_c = Ident_label)) + 
  geom_line(size = .7, aes(linetype = Condition)) +
  xlab("MFQ responses for self") + 
  ylab("MFQ responses for target") + 
  scale_linetype_manual("Issue polarization", 
                        breaks = c(1,2,3,4), 
                        labels = c ("Political\ncandidate",
                                   "Racial\ndisadvantage",
                                   "Marijuana\nlegalization",
                                   "Peanut\nbutter"),
                        values = c("solid",
                                   "twodash", 
                                   "dotted", 
                                   "dashed")) + 
  ggtitle("Projection predicted by identification and issue polarization") +
  theme_minimal(base_size = 13) +
  theme(legend.text = element_text(size = 10),
        legend.key.width = unit(1, "cm"),
        legend.position = "bottom")

ggpreview(width = 8, height = 6, units = "in")

Interesting! With individualizing morals, people counter project more in the political condition when they are average identified than in the overall model and people strongly counter project in both the political and racial disadvantage conditions when highly identified. The overall model did not show counter projection in the racial disadvantage condition when people were highly identified.

knitr::opts_chunk$set(fig.width = 8, fig.height = 6) 

ggplot(f1_indiv_sum_cond, aes(x = MFQSelf_c, 
                              y = lsmean, 
                              group = as.factor(Condition)))+
  geom_line(size = 1, aes(linetype = Condition)) +
  xlab("MFQ responses for self") + 
  ylab("MFQ responses for target") + 
  scale_linetype_manual("Issue polarization", 
                      breaks = c(1,2,3,4),
                      labels = c("Political\ncandidate",
                                  "Racial\ndisadvantage", 
                                  "Marijuana\nlegalization",
                                  "Peanut\nbutter"),
                       values = c("solid",
                                   "twodash", 
                                   "dotted", 
                                   "dashed")) +
  ggtitle("Effect of issue polarization on projection with out-group targets") +
  theme_minimal(base_size = 13) +
  theme(legend.text = element_text(size = 10),
        legend.key.width = unit(1.2, "cm"))

ggpreview(width = 8, height = 6, units = "in")

Only difference is now there is counter projection in the political candidate condition and the racial disadvantage condition is pretty flat.

knitr::opts_chunk$set(fig.width = 8, fig.height = 6) 

ggplot(f1_indiv_ident, aes(MFQSelf_c, 
                               fit, 
                               group = Identification)) +
  geom_line(aes(linetype = Identification), size = .7) +
  scale_linetype_manual("Identification", 
                       breaks = c(-1.19, 0, 1.19), 
                       labels = c ("Low (1 SD below)", 
                                  "Average ", 
                                  "High (1 SD above)"),
                        values = c("solid",
                                   "twodash", 
                                   "dotted")) +
  theme_minimal(base_size = 13) +
  labs(title = "Projection predicted by degree of in-group identification",
       x = "MFQ responses for self",
       y = "MFQ responses for target")

ggpreview(width = 8, height = 6, units = "in")

Pretty similar to the original.

BFI sub sreas: Follow-up 2

Did this previously in another document, which I have uploaded to this project.

Political ideology breakdown of counter projection

Initial Study - already did this above with who they voted for

Follow-up 1

Look at whether or not they agreed or disagreed with the issue AND if the issue is generally liberal or conservative.

Condition: What issue participants were given; 1 = political candidate, 2 = racial discrimination, 3 = weed, 4 = PB

Issue_agreement: 1 for condition 1 is Clinton, 2 is Trump, for other three conditions 1 is yes to agreeing with issue, 4 is no to agreeing with issue (AKA ignoring PB, 1 = liberal, 4 = conservative beliefs)

f1_ideology_model <- lmer(MFQTarg ~ MFQSelf_c*Identification_c*Condition*Issue_agreement + 
                              (MFQSelf_c|Subject), 
                            data = pptnpf1_comb, na.action = na.omit)

tab_model(f1_ideology_model,
          title = "Projection predicted by ideology")

Projection predicted by ideology

	MFQ Targ
Predictors	Estimates	CI	p
(Intercept)	4.39	4.23 – 4.55	<0.001
MFQSelf_c	0.01	-0.08 – 0.11	0.764
Identification_c	0.30	0.15 – 0.44	<0.001
ConditionPRWvPB	-0.06	-0.17 – 0.06	0.325
ConditionPRvW	0.26	0.14 – 0.39	<0.001
ConditionPvR	-0.10	-0.33 – 0.13	0.417
Issue_agreement	-0.13	-0.22 – -0.05	0.003
MFQSelf_c * Identification_c	-0.14	-0.23 – -0.05	0.002
MFQSelf_c : ConditionPRWvPB	0.16	0.09 – 0.23	<0.001
MFQSelf_c : ConditionPRvW	0.13	0.06 – 0.21	0.001
MFQSelf_c : ConditionPvR	0.29	0.15 – 0.43	<0.001
Identification_c : ConditionPRWvPB	-0.14	-0.23 – -0.05	0.002
Identification_c : ConditionPRvW	-0.12	-0.23 – -0.01	0.038
Identification_c : ConditionPvR	0.13	-0.09 – 0.34	0.247
MFQSelf_c * Issue_agreement	0.07	0.01 – 0.12	0.016
Identification_c * Issue_agreement	-0.15	-0.22 – -0.07	<0.001
ConditionPRWvPB * Issue_agreement	0.05	0.01 – 0.10	0.015
ConditionPRvW * Issue_agreement	-0.10	-0.18 – -0.03	0.006
ConditionPvR * Issue_agreement	0.04	-0.07 – 0.15	0.506
MFQSelf_c : Identification_c : ConditionPRWvPB	0.04	-0.01 – 0.09	0.098
MFQSelf_c : Identification_c : ConditionPRvW	0.08	0.01 – 0.15	0.025
MFQSelf_c : Identification_c : ConditionPvR	-0.00	-0.14 – 0.13	0.943
(MFQSelf_c Identification_c) Issue_agreement	0.03	-0.02 – 0.08	0.187
(MFQSelf_c ConditionPRWvPB) Issue_agreement	-0.04	-0.07 – -0.01	0.002
(MFQSelf_c ConditionPRvW) Issue_agreement	0.00	-0.04 – 0.05	0.855
(MFQSelf_c ConditionPvR) Issue_agreement	-0.06	-0.13 – 0.02	0.139
(Identification_c ConditionPRWvPB) Issue_agreement	0.07	0.03 – 0.10	<0.001
(Identification_c ConditionPRvW) Issue_agreement	0.04	-0.02 – 0.10	0.217
(Identification_c ConditionPvR) Issue_agreement	0.09	-0.01 – 0.20	0.086
(MFQSelf_c Identification_c ConditionPRWvPB) * Issue_agreement	-0.01	-0.03 – 0.01	0.336
(MFQSelf_c Identification_c ConditionPRvW) * Issue_agreement	-0.01	-0.05 – 0.03	0.580
(MFQSelf_c Identification_c ConditionPvR) * Issue_agreement	0.02	-0.05 – 0.08	0.626
Random Effects
σ2	1.32
τ00 Subject	0.38
τ11 Subject.MFQSelf_c	0.12
ρ01 Subject	0.04
ICC	0.33
N Subject	306
Observations	6115
Marginal R2 / Conditional R2	0.132 / 0.423

describe(pptnpf1_comb$Identification_c)

##    vars    n  mean   sd median trimmed  mad   min  max range  skew kurtosis
## X1    1 6115 -0.01 1.19  -0.03       0 1.48 -2.03 1.97     4 -0.22    -0.72
##      se
## X1 0.02

effects_polident_f1 <- effect(term="MFQSelf_c:Identification_c:Condition", 
                        xlevels= list(Identification_c=c(-1.19, 0, 1.19), 
                                      Condition=c("1",
                                                  "2",
                                                  "3",
                                                  "4")), 
                        mod=f1_ideology_model)

effects_polident_f1 <- as.data.frame(effects_polident_f1)

effects_polident_f1$Identification<-as.factor(effects_polident_f1$Identification_c)

f1_polident_ident_label <- c("-1.19"="Low Identification \n(1 SD Below)", 
                       "0" = "Average \nIdentification",
                       "1.19"="High Identification \n(1 SD Above)")

ggplot(effects_polident_f1, aes(MFQSelf_c, fit)) +
  geom_line(size = 1, aes(linetype = Condition)) +
  facet_wrap(~Identification, 
             labeller = labeller(Identification=f1_polident_ident_label)) +
  scale_linetype_manual("Issue polarization", 
                        breaks = c(1,2,3,4), 
                        labels = c ("Political\ncandidate",
                                   "Racial\ndisadvantage",
                                   "Marijuana\nlegalization",
                                   "Peanut\nbutter"),
                        values = c("solid",
                                   "twodash", 
                                   "dotted", 
                                   "dashed")) +
    theme_minimal(base_size = 13) +
  theme(legend.text = element_text(size = 10),
        legend.key.width = unit(1.2, "cm")) +
  labs(title = "Projection predicted by in-group identification and condition",
       subtitle = "Included issue agreement in this model",
       x = "MFQ ratings for self",
       y = "MFQ ratings for target")

describe(pptnpf1_comb$Issue_agreement)

##    vars    n mean   sd median trimmed mad min max range skew kurtosis   se
## X1    1 6115 1.84 1.35      1    1.67   0   1   4     3 0.99    -1.03 0.02

effects_polident_f1_plot2 <- effect(term="MFQSelf_c:Condition:Issue_agreement", 
                        xlevels= list(Issue_agreement=c(1, 4), 
                                      Condition=c("1",
                                                  "2",
                                                  "3",
                                                  "4")), 
                        mod=f1_ideology_model)

effects_polident_f1_plot2 <- as.data.frame(effects_polident_f1_plot2)

effects_polident_f1_plot2$Agreement<-as.factor(effects_polident_f1_plot2$Issue_agreement)

f1_polident_ident_label2 <- c("1" = "Support Clinton, Acknowledge racism, \n Weed legalization, and Smooth PB",
                             "4" = "Support Trump, Did not acknowledge racism, \n disagree with weed legalization, like crunchy peanut butter")

ggplot(effects_polident_f1_plot2, aes(MFQSelf_c, fit)) +
  geom_line(size = 1, aes(linetype = Condition)) +
  facet_wrap(~Agreement, 
             labeller = labeller(Agreement=f1_polident_ident_label2)) +
  scale_linetype_manual("Issue polarization", 
                        breaks = c(1,2,3,4), 
                        labels = c ("Political\ncandidate",
                                   "Racial\ndisadvantage",
                                   "Marijuana\nlegalization",
                                   "Peanut\nbutter"),
                        values = c("solid",
                                   "twodash", 
                                   "dotted", 
                                   "dashed")) +
    theme_minimal(base_size = 13) +
  theme(legend.text = element_text(size = 10),
        legend.key.width = unit(1.2, "cm")) +
  labs(title = "Projection predicted by  issue condition and issue agreement",
       x = "MFQ ratings for self",
       y = "MFQ ratings for target")

Follow-up 2

• opin : Whether participants agreed or disgareed with the issue they were presented (e.g., Do you support marijuana legalization?)
  • 1 = Yes
  • 4 = No

contrasts(npf2_clean$condition) <- "contr.helmert"

f2_ideology_model <- lmer(targ ~ self_c*ident_c*condition*opin + 
                              (self_c|sub_id), 
                            data = npf2_clean, na.action = na.omit)

tab_model(f2_ideology_model,
          title = "Projection predicted by ideology")

Projection predicted by ideology

	targ
Predictors	Estimates	CI	p
(Intercept)	3.14	3.07 – 3.21	<0.001
self_c	0.03	-0.03 – 0.09	0.280
ident_c	-0.07	-0.15 – 0.00	0.061
condition1	0.02	-0.08 – 0.12	0.728
condition2	0.09	0.04 – 0.15	0.001
condition3	-0.00	-0.04 – 0.03	0.849
opin	0.03	-0.00 – 0.07	0.060
self_c * ident_c	-0.11	-0.19 – -0.04	0.001
self_c : condition1	-0.05	-0.13 – 0.04	0.308
self_c : condition2	0.04	-0.01 – 0.09	0.161
self_c : condition3	-0.01	-0.04 – 0.02	0.575
ident_c : condition1	-0.01	-0.13 – 0.11	0.882
ident_c : condition2	0.06	-0.00 – 0.12	0.064
ident_c : condition3	0.01	-0.03 – 0.06	0.481
self_c * opin	-0.01	-0.05 – 0.02	0.387
ident_c * opin	-0.00	-0.05 – 0.04	0.912
condition1 * opin	0.01	-0.04 – 0.05	0.827
condition2 * opin	-0.02	-0.05 – 0.01	0.210
condition3 * opin	-0.00	-0.02 – 0.02	0.684
self_c : ident_c : condition1	0.08	-0.03 – 0.19	0.144
self_c : ident_c : condition2	0.05	-0.01 – 0.11	0.110
self_c : ident_c : condition3	0.01	-0.03 – 0.05	0.545
(self_c * ident_c) * opin	0.04	-0.01 – 0.08	0.089
(self_c * condition1) * opin	0.01	-0.03 – 0.05	0.663
(self_c * condition2) * opin	-0.00	-0.03 – 0.03	0.881
(self_c * condition3) * opin	0.01	-0.01 – 0.03	0.228
(ident_c * condition1) * opin	-0.00	-0.07 – 0.07	0.960
(ident_c * condition2) * opin	-0.04	-0.08 – -0.00	0.032
(ident_c * condition3) * opin	0.00	-0.02 – 0.02	0.994
(self_c * ident_c condition1) opin	-0.03	-0.09 – 0.02	0.254
(self_c * ident_c condition2) opin	-0.04	-0.08 – 0.00	0.062
(self_c * ident_c condition3) opin	-0.00	-0.02 – 0.02	0.808
Random Effects
σ2	1.03
τ00 sub_id	0.05
τ11 sub_id.self_c	0.04
ρ01 sub_id	0.53
ICC	0.10
N sub_id	334
Observations	6324
Marginal R2 / Conditional R2	0.025 / 0.125

Model failed to converge, so results are unreliable. Any plot using intercepts and slopes from this model would also be unreliable.

Demographic differences in counter projection

Initial Study - Race

dem_analysis <- clean_data2 %>% 
  select(subid, MFQ_Number, MFQSelf_c, MFQStereo_c, MFQTarg, ident_c, PT_effects, 
         inoutgroup_effects, par_cand, Race) %>%
  unique() %>% 
  na.omit()

projmodel_stereo_race <- lmer(MFQTarg ~ MFQSelf_c * MFQStereo_c *
                                    inoutgroup_effects * ident_c * Race +
                                    (MFQSelf_c|subid), 
                                 data = dem_analysis)

tab_model(projmodel_stereo_race,
          title = "Projection moderated by stereotyping and race")

Projection moderated by stereotyping and race

	MFQ Targ
Predictors	Estimates	CI	p
(Intercept)	3.25	2.39 – 4.11	<0.001
MFQSelf_c	0.60	0.18 – 1.02	0.005
MFQStereo_c	-0.47	-0.72 – -0.21	<0.001
inoutgroup_effects	1.52	-0.12 – 3.15	0.069
ident_c	0.04	-0.50 – 0.59	0.883
Race [Asian]	0.79	-0.10 – 1.68	0.083
Race [Black]	0.96	0.07 – 1.85	0.034
Race [Latinx]	1.12	0.23 – 2.02	0.014
Race [Middle Eastern]	-0.21	-2.10 – 1.68	0.826
Race [Caucasian]	0.95	0.09 – 1.81	0.031
Race [Multiracial]	1.19	0.28 – 2.11	0.010
Race [Other]	0.70	-0.64 – 2.04	0.309
MFQSelf_c * MFQStereo_c	0.08	-0.03 – 0.20	0.154
MFQSelf_c * inoutgroup_effects	0.20	-0.61 – 1.00	0.629
MFQStereo_c * inoutgroup_effects	1.10	0.58 – 1.62	<0.001
MFQSelf_c * ident_c	-0.15	-0.45 – 0.14	0.300
MFQStereo_c * ident_c	0.09	-0.16 – 0.33	0.489
inoutgroup_effects * ident_c	0.32	-0.02 – 0.66	0.064
MFQSelf_c * Race [Asian]	-0.53	-0.97 – -0.10	0.016
MFQSelf_c * Race [Black]	-0.48	-0.91 – -0.05	0.030
MFQSelf_c * Race [Latinx]	-0.44	-0.87 – -0.00	0.048
MFQSelf_c * Race [Middle Eastern]	-1.02	-2.20 – 0.17	0.093
MFQSelf_c * Race [Caucasian]	-0.45	-0.87 – -0.03	0.034
MFQSelf_c * Race [Multiracial]	-0.52	-0.96 – -0.07	0.023
MFQSelf_c * Race [Other]	-0.36	-1.09 – 0.37	0.334
MFQStereo_c * Race [Asian]	0.63	0.36 – 0.89	<0.001
MFQStereo_c * Race [Black]	0.54	0.28 – 0.80	<0.001
MFQStereo_c * Race [Latinx]	0.50	0.23 – 0.76	<0.001
MFQStereo_c * Race [Middle Eastern]	0.12	-0.74 – 0.97	0.790
MFQStereo_c * Race [Caucasian]	0.54	0.29 – 0.80	<0.001
MFQStereo_c * Race [Multiracial]	0.60	0.32 – 0.87	<0.001
MFQStereo_c * Race [Other]	-0.09	-0.63 – 0.45	0.754
inoutgroup_effects * Race [Asian]	-1.14	-2.85 – 0.56	0.188
inoutgroup_effects * Race [Black]	-0.98	-2.68 – 0.72	0.257
inoutgroup_effects * Race [Latinx]	-1.11	-2.82 – 0.60	0.203
inoutgroup_effects * Race [Caucasian]	-1.32	-2.96 – 0.33	0.116
inoutgroup_effects * Race [Multiracial]	-1.20	-3.00 – 0.61	0.193
ident_c * Race [Asian]	-0.00	-0.57 – 0.56	0.987
ident_c * Race [Black]	-0.06	-0.63 – 0.51	0.834
ident_c * Race [Latinx]	0.04	-0.52 – 0.60	0.893
ident_c * Race [Caucasian]	-0.01	-0.55 – 0.54	0.981
ident_c * Race [Multiracial]	0.05	-0.55 – 0.65	0.869
ident_c * Race [Other]	0.21	-0.44 – 0.87	0.521
(MFQSelf_c * MFQStereo_c) * inoutgroup_effects	-0.18	-0.43 – 0.06	0.147
(MFQSelf_c * MFQStereo_c) * ident_c	-0.03	-0.15 – 0.08	0.576
(MFQSelf_c inoutgroup_effects) ident_c	0.05	-0.11 – 0.21	0.563
(MFQStereo_c inoutgroup_effects) ident_c	-0.02	-0.11 – 0.07	0.670
(MFQSelf_c * MFQStereo_c) * Race [Asian]	-0.05	-0.18 – 0.07	0.388
(MFQSelf_c * MFQStereo_c) * Race [Black]	-0.13	-0.25 – -0.01	0.038
(MFQSelf_c * MFQStereo_c) * Race [Latinx]	-0.10	-0.22 – 0.02	0.117
(MFQSelf_c * MFQStereo_c) * Race [Middle Eastern]	-0.14	-1.63 – 1.35	0.855
(MFQSelf_c * MFQStereo_c) * Race [Caucasian]	-0.11	-0.23 – 0.00	0.060
(MFQSelf_c * MFQStereo_c) * Race [Multiracial]	-0.19	-0.31 – -0.06	0.003
(MFQSelf_c * MFQStereo_c) * Race [Other]	-0.14	-0.37 – 0.09	0.231
(MFQSelf_c inoutgroup_effects) Race [Asian]	0.12	-0.72 – 0.95	0.786
(MFQSelf_c inoutgroup_effects) Race [Black]	0.09	-0.75 – 0.93	0.836
(MFQSelf_c inoutgroup_effects) Race [Latinx]	0.13	-0.71 – 0.97	0.770
(MFQSelf_c inoutgroup_effects) Race [Caucasian]	0.17	-0.63 – 0.98	0.672
(MFQSelf_c inoutgroup_effects) Race [Multiracial]	0.11	-0.77 – 1.00	0.807
(MFQStereo_c inoutgroup_effects) Race [Asian]	-0.46	-0.99 – 0.08	0.093
(MFQStereo_c inoutgroup_effects) Race [Black]	-0.68	-1.21 – -0.15	0.012
(MFQStereo_c inoutgroup_effects) Race [Latinx]	-0.68	-1.22 – -0.15	0.012
(MFQStereo_c inoutgroup_effects) Race [Caucasian]	-0.48	-0.99 – 0.04	0.071
(MFQStereo_c inoutgroup_effects) Race [Multiracial]	-0.52	-1.09 – 0.05	0.073
(MFQSelf_c * ident_c) * Race [Asian]	0.18	-0.12 – 0.48	0.249
(MFQSelf_c * ident_c) * Race [Black]	0.14	-0.17 – 0.44	0.378
(MFQSelf_c * ident_c) * Race [Latinx]	0.24	-0.06 – 0.54	0.119
(MFQSelf_c * ident_c) * Race [Caucasian]	0.16	-0.14 – 0.45	0.295
(MFQSelf_c * ident_c) * Race [Multiracial]	0.13	-0.19 – 0.44	0.434
(MFQSelf_c * ident_c) * Race [Other]	0.17	-0.18 – 0.53	0.344
(MFQStereo_c * ident_c) * Race [Asian]	-0.15	-0.40 – 0.10	0.231
(MFQStereo_c * ident_c) * Race [Black]	-0.16	-0.41 – 0.09	0.208
(MFQStereo_c * ident_c) * Race [Latinx]	-0.05	-0.30 – 0.20	0.708
(MFQStereo_c * ident_c) * Race [Caucasian]	-0.13	-0.37 – 0.12	0.318
(MFQStereo_c * ident_c) * Race [Multiracial]	-0.10	-0.35 – 0.15	0.438
(MFQStereo_c * ident_c) * Race [Other]	-0.19	-0.48 – 0.10	0.198
(inoutgroup_effects ident_c) Race [Asian]	-0.10	-0.54 – 0.33	0.643
(inoutgroup_effects ident_c) Race [Black]	-0.26	-0.73 – 0.20	0.270
(inoutgroup_effects ident_c) Race [Latinx]	0.12	-0.31 – 0.56	0.575
(inoutgroup_effects ident_c) Race [Caucasian]	-0.08	-0.42 – 0.27	0.661
(MFQSelf_c * MFQStereo_c * inoutgroup_effects) * ident_c	-0.04	-0.08 – 0.01	0.113
(MFQSelf_c * MFQStereo_c * inoutgroup_effects) * Race [Asian]	0.19	-0.07 – 0.45	0.159
(MFQSelf_c * MFQStereo_c * inoutgroup_effects) * Race [Black]	0.19	-0.07 – 0.45	0.146
(MFQSelf_c * MFQStereo_c * inoutgroup_effects) * Race [Latinx]	0.19	-0.06 – 0.45	0.138
(MFQSelf_c * MFQStereo_c * inoutgroup_effects) * Race [Caucasian]	0.19	-0.05 – 0.44	0.123
(MFQSelf_c * MFQStereo_c * inoutgroup_effects) * Race [Multiracial]	0.06	-0.21 – 0.33	0.659
(MFQSelf_c * MFQStereo_c * ident_c) * Race [Asian]	0.00	-0.12 – 0.12	0.996
(MFQSelf_c * MFQStereo_c * ident_c) * Race [Black]	0.02	-0.10 – 0.14	0.742
(MFQSelf_c * MFQStereo_c * ident_c) * Race [Latinx]	0.04	-0.07 – 0.16	0.455
(MFQSelf_c * MFQStereo_c * ident_c) * Race [Caucasian]	0.03	-0.09 – 0.14	0.656
(MFQSelf_c * MFQStereo_c * ident_c) * Race [Multiracial]	0.02	-0.10 – 0.14	0.798
(MFQSelf_c * MFQStereo_c * ident_c) * Race [Other]	0.00	-0.13 – 0.13	0.979
(MFQSelf_c inoutgroup_effects ident_c) * Race [Asian]	-0.04	-0.25 – 0.17	0.687
(MFQSelf_c inoutgroup_effects ident_c) * Race [Black]	-0.02	-0.25 – 0.22	0.898
(MFQSelf_c inoutgroup_effects ident_c) * Race [Latinx]	-0.15	-0.36 – 0.05	0.138
(MFQSelf_c inoutgroup_effects ident_c) * Race [Caucasian]	-0.04	-0.20 – 0.12	0.617
(MFQStereo_c inoutgroup_effects ident_c) * Race [Asian]	0.13	0.01 – 0.24	0.028
(MFQStereo_c inoutgroup_effects ident_c) * Race [Black]	-0.08	-0.22 – 0.05	0.226
(MFQStereo_c inoutgroup_effects ident_c) * Race [Latinx]	0.05	-0.06 – 0.16	0.379
(MFQStereo_c inoutgroup_effects ident_c) * Race [Caucasian]	0.03	-0.07 – 0.12	0.588
(MFQSelf_c * MFQStereo_c * inoutgroup_effects ident_c) Race [Asian]	0.11	0.05 – 0.18	0.001
(MFQSelf_c * MFQStereo_c * inoutgroup_effects ident_c) Race [Black]	0.04	-0.03 – 0.12	0.270
(MFQSelf_c * MFQStereo_c * inoutgroup_effects ident_c) Race [Latinx]	0.05	-0.00 – 0.11	0.073
(MFQSelf_c * MFQStereo_c * inoutgroup_effects ident_c) Race [Caucasian]	0.03	-0.01 – 0.08	0.170
Random Effects
σ2	1.30
τ00 subid	0.46
τ11 subid.MFQSelf_c	0.09
ρ01 subid	-0.02
ICC	0.33
N subid	670
Observations	19925
Marginal R2 / Conditional R2	0.226 / 0.482

Model is rank deficient, not enough power to check if results differ by race. Could possibly do white vs minorities, but that doesn’t feel telling enough.

Initial Study - Gender

gender_analysis <- clean_data2 %>% 
  select(subid, MFQ_Number, MFQSelf_c, MFQStereo_c, MFQTarg, ident_c, PT_effects, 
         inoutgroup_effects, par_cand, Gender) %>%
  unique() %>% 
  na.omit()

projmodel_stereo_gender <- lmer(MFQTarg ~ MFQSelf_c * MFQStereo_c *
                                    inoutgroup_effects * ident_c * Gender +
                                    (MFQSelf_c|subid), 
                                 data = gender_analysis)

tab_model(projmodel_stereo_gender,
          title = "Projection moderated by stereotyping and gender")

Projection moderated by stereotyping and gender

	MFQ Targ
Predictors	Estimates	CI	p
(Intercept)	4.03	3.94 – 4.12	<0.001
MFQSelf_c	0.14	0.09 – 0.19	<0.001
MFQStereo_c	0.04	0.01 – 0.07	0.003
inoutgroup_effects	0.30	0.11 – 0.48	0.002
ident_c	0.03	-0.03 – 0.08	0.322
Gender [Female]	0.25	0.14 – 0.37	<0.001
Gender [Other]	0.14	-0.91 – 1.19	0.792
MFQSelf_c * MFQStereo_c	-0.03	-0.04 – -0.01	<0.001
MFQSelf_c * inoutgroup_effects	0.35	0.26 – 0.44	<0.001
MFQStereo_c * inoutgroup_effects	0.60	0.55 – 0.65	<0.001
MFQSelf_c * ident_c	0.00	-0.02 – 0.03	0.727
MFQStereo_c * ident_c	-0.05	-0.07 – -0.04	<0.001
inoutgroup_effects * ident_c	0.30	0.19 – 0.40	<0.001
MFQSelf_c * Gender [Female]	-0.00	-0.06 – 0.05	0.954
MFQSelf_c * Gender [Other]	0.08	-0.41 – 0.57	0.756
MFQStereo_c * Gender [Female]	0.05	0.02 – 0.09	0.002
MFQStereo_c * Gender [Other]	0.04	-0.31 – 0.40	0.824
inoutgroup_effects * Gender [Female]	-0.10	-0.32 – 0.13	0.407
inoutgroup_effects * Gender [Other]	0.43	-1.67 – 2.53	0.689
ident_c * Gender [Female]	0.02	-0.04 – 0.09	0.493
ident_c * Gender [Other]	-0.15	-0.56 – 0.25	0.452
(MFQSelf_c * MFQStereo_c) * inoutgroup_effects	0.02	-0.01 – 0.05	0.190
(MFQSelf_c * MFQStereo_c) * ident_c	-0.01	-0.01 – 0.00	0.153
(MFQSelf_c inoutgroup_effects) ident_c	0.03	-0.02 – 0.08	0.253
(MFQStereo_c inoutgroup_effects) ident_c	0.02	-0.01 – 0.05	0.202
(MFQSelf_c * MFQStereo_c) * Gender [Female]	0.00	-0.02 – 0.02	0.964
(MFQSelf_c * MFQStereo_c) * Gender [Other]	0.00	-0.17 – 0.17	0.995
(MFQSelf_c inoutgroup_effects) Gender [Female]	0.04	-0.08 – 0.15	0.536
(MFQSelf_c inoutgroup_effects) Gender [Other]	-0.45	-1.43 – 0.53	0.366
(MFQStereo_c inoutgroup_effects) Gender [Female]	-0.00	-0.07 – 0.06	0.949
(MFQStereo_c inoutgroup_effects) Gender [Other]	-0.23	-0.94 – 0.48	0.530
(MFQSelf_c * ident_c) * Gender [Female]	0.00	-0.03 – 0.03	0.914
(MFQSelf_c * ident_c) * Gender [Other]	-0.01	-0.19 – 0.18	0.953
(MFQStereo_c * ident_c) * Gender [Female]	0.03	0.01 – 0.05	0.001
(MFQStereo_c * ident_c) * Gender [Other]	0.04	-0.14 – 0.22	0.650
(inoutgroup_effects ident_c) Gender [Female]	-0.09	-0.22 – 0.04	0.163
(inoutgroup_effects ident_c) Gender [Other]	0.12	-0.68 – 0.93	0.768
(MFQSelf_c * MFQStereo_c * inoutgroup_effects) * ident_c	-0.01	-0.03 – 0.00	0.138
(MFQSelf_c * MFQStereo_c * inoutgroup_effects) * Gender [Female]	-0.01	-0.05 – 0.02	0.518
(MFQSelf_c * MFQStereo_c * inoutgroup_effects) * Gender [Other]	-0.27	-0.61 – 0.08	0.127
(MFQSelf_c * MFQStereo_c * ident_c) * Gender [Female]	-0.00	-0.01 – 0.01	0.766
(MFQSelf_c * MFQStereo_c * ident_c) * Gender [Other]	0.01	-0.08 – 0.09	0.900
(MFQSelf_c inoutgroup_effects ident_c) * Gender [Female]	-0.03	-0.10 – 0.03	0.282
(MFQSelf_c inoutgroup_effects ident_c) * Gender [Other]	-0.29	-0.66 – 0.09	0.132
(MFQStereo_c inoutgroup_effects ident_c) * Gender [Female]	-0.02	-0.06 – 0.02	0.261
(MFQStereo_c inoutgroup_effects ident_c) * Gender [Other]	0.22	-0.13 – 0.58	0.218
(MFQSelf_c * MFQStereo_c * inoutgroup_effects ident_c) Gender [Female]	0.02	-0.00 – 0.04	0.065
(MFQSelf_c * MFQStereo_c * inoutgroup_effects ident_c) Gender [Other]	-0.03	-0.20 – 0.14	0.754
Random Effects
σ2	1.31
τ00 subid	0.44
τ11 subid.MFQSelf_c	0.08
ρ01 subid	-0.04
ICC	0.32
N subid	673
Observations	20015
Marginal R2 / Conditional R2	0.223 / 0.473

Gender does not moderate the results.

Follow-up 1 - Gender

f1_gender <- lmer(MFQTarg ~ MFQSelf_c*Identification_c*Condition*Gender + 
                              (MFQSelf_c|Subject), 
                            data = pptnpf1_comb, na.action = na.omit)

When adding gender,the model failed to converge.

Follow-up 1 - Race

f1_race <- lmer(MFQTarg ~ MFQSelf_c*Identification_c*Condition*Race + 
                              (MFQSelf_c|Subject), 
                            data = pptnpf1_comb, na.action = na.omit)

When adding race, the fixed effect model was rank deficient and dropped predictors/comparisons.

Follow-up 2 - Gender

npf2_clean <- npf2_clean %>% 
  mutate(gender = as.factor(gender),
         race = as.factor(race))

f2_gender <- lmer(targ ~ self_c*ident_c*condition*gender + 
                              (self_c|sub_id), 
                            data = npf2_clean, na.action = na.omit)

Fixed effect model is rank deficient

Follow-up 2 - Race

f2_race <- lmer(targ ~ self_c*ident_c*condition*race + 
                              (self_c|sub_id), 
                            data = npf2_clean, na.action = na.omit)

Fixed effect model is rank deficient. We don’t have enough power to include extra categorical variables…

Downstream outcomes impacted by counter projection - Initial Study

downstream_analysis <- clean_data2 %>% 
  select(subid, MFQ_Number, MFQSelf, MFQSelf_c, MFQStereo_c, MFQTarg, ident_c,
         inoutgroup_effects, par_cand, Liking, WilltoDiscuss, Contentious, Treatment, Valid, MFQStereo) %>%
  unique() %>% 
  na.omit() %>% 
  group_by(subid) %>% 
  #getting the difference scores
  #decided to not mean center difference scores, since 0 is meaningful (no difference = 100% projection) 
  mutate(proj_diff = abs(mean(MFQTarg) - mean(MFQSelf)),
         stereo_diff = abs(mean(MFQTarg)- mean(MFQStereo)))

mfq_binding_diff <- mfq_binding %>% 
  group_by(subid) %>% 
  mutate(proj_bind_diff = abs(mean(MFQTarg) - mean(MFQSelf)),
         stereo_bind_diff = abs(mean(MFQTarg) - mean(MFQStereo)))

mfq_individual_diff <- mfq_individual %>% 
 group_by(subid) %>% 
  mutate(proj_indiv_diff = abs(mean(MFQTarg) - mean(MFQStereo)),
         stereo_indiv_diff = abs(mean(MFQTarg) - mean(MFQStereo))) %>% 
  na.omit()

How much do you like the target?

Overall model

liking_overall <- lm(Liking ~ proj_diff*stereo_diff*inoutgroup_effects*ident_c, downstream_analysis)
summary(liking_overall)

## 
## Call:
## lm(formula = Liking ~ proj_diff * stereo_diff * inoutgroup_effects * 
##     ident_c, data = downstream_analysis)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -3.4540 -0.8020  0.2304  0.7574  4.0435 
## 
## Coefficients:
##                                                   Estimate Std. Error t value
## (Intercept)                                       4.719954   0.021130 223.377
## proj_diff                                        -0.190032   0.034956  -5.436
## stereo_diff                                      -0.165177   0.035573  -4.643
## inoutgroup_effects                                1.663988   0.042260  39.375
## ident_c                                           0.153919   0.011369  13.539
## proj_diff:stereo_diff                             0.077122   0.045368   1.700
## proj_diff:inoutgroup_effects                     -0.123782   0.069911  -1.771
## stereo_diff:inoutgroup_effects                    0.630144   0.071146   8.857
## proj_diff:ident_c                                -0.111098   0.015032  -7.391
## stereo_diff:ident_c                               0.106644   0.018233   5.849
## inoutgroup_effects:ident_c                        0.230630   0.022737  10.143
## proj_diff:stereo_diff:inoutgroup_effects         -0.105041   0.090736  -1.158
## proj_diff:stereo_diff:ident_c                    -0.017069   0.014762  -1.156
## proj_diff:inoutgroup_effects:ident_c             -0.021255   0.030064  -0.707
## stereo_diff:inoutgroup_effects:ident_c            0.165531   0.036467   4.539
## proj_diff:stereo_diff:inoutgroup_effects:ident_c  0.003271   0.029524   0.111
##                                                              Pr(>|t|)    
## (Intercept)                                      < 0.0000000000000002 ***
## proj_diff                                           0.000000055011895 ***
## stereo_diff                                         0.000003450972714 ***
## inoutgroup_effects                               < 0.0000000000000002 ***
## ident_c                                          < 0.0000000000000002 ***
## proj_diff:stereo_diff                                          0.0892 .  
## proj_diff:inoutgroup_effects                                   0.0766 .  
## stereo_diff:inoutgroup_effects                   < 0.0000000000000002 ***
## proj_diff:ident_c                                   0.000000000000152 ***
## stereo_diff:ident_c                                 0.000000005027905 ***
## inoutgroup_effects:ident_c                       < 0.0000000000000002 ***
## proj_diff:stereo_diff:inoutgroup_effects                       0.2470    
## proj_diff:stereo_diff:ident_c                                  0.2476    
## proj_diff:inoutgroup_effects:ident_c                           0.4796    
## stereo_diff:inoutgroup_effects:ident_c              0.000005678979911 ***
## proj_diff:stereo_diff:inoutgroup_effects:ident_c               0.9118    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.262 on 19882 degrees of freedom
## Multiple R-squared:  0.4223, Adjusted R-squared:  0.4219 
## F-statistic: 969.1 on 15 and 19882 DF,  p-value: < 0.00000000000000022

Overall model Plots

#main effect of projection
ggplot(downstream_analysis, aes(proj_diff, Liking)) +
  geom_smooth(method = "lm") +
  theme_minimal() +
  labs(title = "Liking predicted by projection of moral values",
       subtitle = "Projection is the absolute value of the difference score \nbetween average MFQ scores for target and self",
       x = "Projection (Higher numbers indicate lower projection)",
       y = "Liking")

Counter projection can be seen as anything with a distance over 1, as that means that people are not listing their score as the same as the targets. The greater the distance demonstrates the greater the counter projection. As can be seen, as people counter project more, they report liking the target less. People who project normally report liking the target the most.

#interaction between projection and degree of identification
describe(downstream_analysis$ident_c)

##    vars     n  mean   sd median trimmed mad   min  max range  skew kurtosis
## X1    1 19898 -0.03 1.78  -0.01   -0.04 1.8 -4.23 3.77     8 -0.02    -0.32
##      se
## X1 0.01

liking_overall_2way <- effect(term = "proj_diff:ident_c", 
                           xlevels = list(ident_c = c(-1.78, 0, 1.78)), 
                           mod = liking_overall)

liking_overall_2way <- as.data.frame(liking_overall_2way)

liking_overall_2way$Identification <- as.factor(liking_overall_2way$ident_c)

ggplot(liking_overall_2way, aes(proj_diff, fit, group = Identification, color = Identification)) +
  geom_smooth(method = "lm") +
  theme_minimal() +
  scale_color_discrete("Identification", breaks = c(-1.78, 0, 1.78), 
                       labels = c("Weak identification \n (1 SD below)", 
                                  "Average identification", 
                                  "High identification \n (1 SD above)")) +
  labs(title = "Liking predicted by projection of moral values",
       subtitle = "Projection is the absolute value of the difference score between average \nMFQ scores for target and self. Higher scores = greater counter projection.",
       x = "Projection (Higher = more counter projection)",
       y = "Liking")

Those who weakly identify continue to like the target, even as the distance between their scores on the MFQ and their target’s increase (becoming less similar, less projection). As in-group identification increases, self-reported liking of the target decreases as counter projection increases.

Binding model

liking_bind <- lm(Liking ~ proj_bind_diff*stereo_bind_diff*inoutgroup_effects*ident_c, mfq_binding_diff)
summary(liking_bind)

## 
## Call:
## lm(formula = Liking ~ proj_bind_diff * stereo_bind_diff * inoutgroup_effects * 
##     ident_c, data = mfq_binding_diff)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -3.5388 -0.7824  0.1056  0.7848  3.5943 
## 
## Coefficients:
##                                                             Estimate Std. Error
## (Intercept)                                                 4.794162   0.026896
## proj_bind_diff                                             -0.090386   0.029169
## stereo_bind_diff                                            0.028208   0.030557
## inoutgroup_effects                                          1.464714   0.053792
## ident_c                                                     0.089831   0.016064
## proj_bind_diff:stereo_bind_diff                            -0.044362   0.026371
## proj_bind_diff:inoutgroup_effects                           0.139702   0.058339
## stereo_bind_diff:inoutgroup_effects                        -0.042776   0.061114
## proj_bind_diff:ident_c                                     -0.004495   0.013961
## stereo_bind_diff:ident_c                                    0.115828   0.017389
## inoutgroup_effects:ident_c                                  0.190607   0.032128
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects          0.112589   0.052742
## proj_bind_diff:stereo_bind_diff:ident_c                    -0.038439   0.011566
## proj_bind_diff:inoutgroup_effects:ident_c                   0.124375   0.027921
## stereo_bind_diff:inoutgroup_effects:ident_c                 0.129997   0.034779
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects:ident_c -0.064567   0.023133
##                                                            t value
## (Intercept)                                                178.249
## proj_bind_diff                                              -3.099
## stereo_bind_diff                                             0.923
## inoutgroup_effects                                          27.229
## ident_c                                                      5.592
## proj_bind_diff:stereo_bind_diff                             -1.682
## proj_bind_diff:inoutgroup_effects                            2.395
## stereo_bind_diff:inoutgroup_effects                         -0.700
## proj_bind_diff:ident_c                                      -0.322
## stereo_bind_diff:ident_c                                     6.661
## inoutgroup_effects:ident_c                                   5.933
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects           2.135
## proj_bind_diff:stereo_bind_diff:ident_c                     -3.323
## proj_bind_diff:inoutgroup_effects:ident_c                    4.455
## stereo_bind_diff:inoutgroup_effects:ident_c                  3.738
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects:ident_c  -2.791
##                                                                        Pr(>|t|)
## (Intercept)                                                < 0.0000000000000002
## proj_bind_diff                                                         0.001949
## stereo_bind_diff                                                       0.355960
## inoutgroup_effects                                         < 0.0000000000000002
## ident_c                                                         0.0000000229220
## proj_bind_diff:stereo_bind_diff                                        0.092554
## proj_bind_diff:inoutgroup_effects                                      0.016651
## stereo_bind_diff:inoutgroup_effects                                    0.483980
## proj_bind_diff:ident_c                                                 0.747462
## stereo_bind_diff:ident_c                                        0.0000000000284
## inoutgroup_effects:ident_c                                      0.0000000030606
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects                     0.032805
## proj_bind_diff:stereo_bind_diff:ident_c                                0.000892
## proj_bind_diff:inoutgroup_effects:ident_c                       0.0000084843620
## stereo_bind_diff:inoutgroup_effects:ident_c                            0.000186
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects:ident_c             0.005261
##                                                               
## (Intercept)                                                ***
## proj_bind_diff                                             ** 
## stereo_bind_diff                                              
## inoutgroup_effects                                         ***
## ident_c                                                    ***
## proj_bind_diff:stereo_bind_diff                            .  
## proj_bind_diff:inoutgroup_effects                          *  
## stereo_bind_diff:inoutgroup_effects                           
## proj_bind_diff:ident_c                                        
## stereo_bind_diff:ident_c                                   ***
## inoutgroup_effects:ident_c                                 ***
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects         *  
## proj_bind_diff:stereo_bind_diff:ident_c                    ***
## proj_bind_diff:inoutgroup_effects:ident_c                  ***
## stereo_bind_diff:inoutgroup_effects:ident_c                ***
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects:ident_c ** 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.215 on 11926 degrees of freedom
## Multiple R-squared:  0.4644, Adjusted R-squared:  0.4637 
## F-statistic: 689.3 on 15 and 11926 DF,  p-value: < 0.00000000000000022

Binding model plots

#main effect of projection
ggplot(mfq_binding_diff, aes(proj_bind_diff, Liking)) +
  geom_smooth(method = "lm") +
  theme_minimal() +
  labs(title = "Liking predicted by projection of binding morals",
       subtitle = "Projection is the absolute value of the difference score \nbetween average MFQ scores for target and self",
       x = "Projection (Higher numbers indicate lower projection)",
       y = "Liking")

Same as overall plot.

liking_bind_2way_group <- effect(term = "proj_bind_diff:inoutgroup_effects", 
                           xlevels = list(inoutgroup_effects = c(-.5,.5)), 
                           mod = liking_bind)

liking_bind_2way_group <- as.data.frame(liking_bind_2way_group)

liking_bind_2way_group$group <- as.factor(liking_bind_2way_group$inoutgroup_effects)

ggplot(liking_bind_2way_group, aes(proj_bind_diff, fit, group = group, color = group)) +
 geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black", 
              aes(linetype = group)) +
  theme_minimal() +
  scale_linetype_manual("Group", breaks = c(-.5, .5), 
                       labels = c("Out-group", "In-group"),
                       values = c("solid", 
                                   "dashed")) +
  labs(title = "Liking predicted by projection of binding moral values",
       subtitle = "Projection is the absolute value of the difference score between average \nMFQ scores for target and self. Higher scores = greater counter projection.",
       x = "Projection (Higher = more counter projection)",
       y = "Liking")

Liking stays high with the in-group, regardless of their degree of projection. Liking is highest for the out-group when they are projecting. As counter projection increases (the distance becomes further), liking decreases for the out-group.

describe(stereo_analysis$ident_c)

##    vars     n  mean   sd median trimmed mad   min  max range  skew kurtosis
## X1    1 20045 -0.03 1.78  -0.01   -0.04 1.8 -4.23 3.77     8 -0.02    -0.32
##      se
## X1 0.01

binding_liking_3way <- effect(term = "proj_bind_diff:inoutgroup_effects:ident_c", 
                           xlevels = list(ident_c = c(-1.78, 0, 1.78), 
                                         inoutgroup_effects = c(-.5,.5)), 
                           mod = liking_bind)

binding_liking_3way <- as.data.frame(binding_liking_3way)

binding_liking_3way$Group <- as.factor(binding_liking_3way$inoutgroup_effects)
binding_liking_3way$Identification <- as.factor(binding_liking_3way$ident_c)

ident_label <- c("-1.78" = "Low identification \n (1 SD below)", 
                 "0" = "Average identification",
                 "1.78" = "High identificatiion \n(1 SD above)")

ggplot(binding_liking_3way, aes(proj_bind_diff,
                             fit, 
                             group = Group,
                             color = Group)) +
  geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black", 
              aes(linetype = Group)) +
  facet_wrap(~Identification, labeller = 
               labeller(Identification = ident_label)) +
  scale_linetype_manual("Group", breaks = c(-.5, .5), 
                       labels = c("Out-group", "In-group"),
                       values = c("solid", 
                                   "dashed")) +
  theme_minimal() +
  labs(title = "Liking predicted by projection of binding morals, \nin-group identification and group membership",
       subtitle = "Controlling for stereotyping",
       x = "Projection difference scores (higher = more counter projection)",
       y = "Liking")

As can be seen in the plot above, people report liking an in-group target at a similiar rate regardless of projection, while liking of the out-group target decreases as both counter projection and in-group identification increase. Thus, the people who report liking the least are considering an out-group target and are highly identified with their in-group as well as projecting their own morals the least.

Individuating model

liking_indiv <- lm(Liking ~ proj_indiv_diff*stereo_indiv_diff*inoutgroup_effects*ident_c, mfq_individual_diff)
summary(liking_indiv)

## 
## Call:
## lm(formula = Liking ~ proj_indiv_diff * stereo_indiv_diff * inoutgroup_effects * 
##     ident_c, data = mfq_individual_diff)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -3.5668 -0.7639  0.1056  0.7996  3.6753 
## 
## Coefficients: (4 not defined because of singularities)
##                                                              Estimate
## (Intercept)                                                   4.87853
## proj_indiv_diff                                              -0.17945
## stereo_indiv_diff                                                  NA
## inoutgroup_effects                                            1.28560
## ident_c                                                       0.04418
## proj_indiv_diff:stereo_indiv_diff                            -0.08643
## proj_indiv_diff:inoutgroup_effects                            0.62506
## stereo_indiv_diff:inoutgroup_effects                               NA
## proj_indiv_diff:ident_c                                       0.25313
## stereo_indiv_diff:ident_c                                          NA
## inoutgroup_effects:ident_c                                    0.17232
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects         -0.12613
## proj_indiv_diff:stereo_indiv_diff:ident_c                    -0.10738
## proj_indiv_diff:inoutgroup_effects:ident_c                    0.47335
## stereo_indiv_diff:inoutgroup_effects:ident_c                       NA
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects:ident_c -0.20190
##                                                              Std. Error t value
## (Intercept)                                                     0.03012 161.976
## proj_indiv_diff                                                 0.06826  -2.629
## stereo_indiv_diff                                                    NA      NA
## inoutgroup_effects                                              0.06024  21.342
## ident_c                                                         0.01492   2.960
## proj_indiv_diff:stereo_indiv_diff                               0.03666  -2.358
## proj_indiv_diff:inoutgroup_effects                              0.13653   4.578
## stereo_indiv_diff:inoutgroup_effects                                 NA      NA
## proj_indiv_diff:ident_c                                         0.03434   7.372
## stereo_indiv_diff:ident_c                                            NA      NA
## inoutgroup_effects:ident_c                                      0.02985   5.773
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects            0.07331  -1.720
## proj_indiv_diff:stereo_indiv_diff:ident_c                       0.01721  -6.238
## proj_indiv_diff:inoutgroup_effects:ident_c                      0.06868   6.892
## stereo_indiv_diff:inoutgroup_effects:ident_c                         NA      NA
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects:ident_c    0.03443  -5.865
##                                                                          Pr(>|t|)
## (Intercept)                                                  < 0.0000000000000002
## proj_indiv_diff                                                           0.00859
## stereo_indiv_diff                                                              NA
## inoutgroup_effects                                           < 0.0000000000000002
## ident_c                                                                   0.00308
## proj_indiv_diff:stereo_indiv_diff                                         0.01841
## proj_indiv_diff:inoutgroup_effects                              0.000004761326363
## stereo_indiv_diff:inoutgroup_effects                                           NA
## proj_indiv_diff:ident_c                                         0.000000000000185
## stereo_indiv_diff:ident_c                                                      NA
## inoutgroup_effects:ident_c                                      0.000000008089039
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects                      0.08539
## proj_indiv_diff:stereo_indiv_diff:ident_c                       0.000000000465508
## proj_indiv_diff:inoutgroup_effects:ident_c                      0.000000000005905
## stereo_indiv_diff:inoutgroup_effects:ident_c                                   NA
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects:ident_c    0.000000004673593
##                                                                 
## (Intercept)                                                  ***
## proj_indiv_diff                                              ** 
## stereo_indiv_diff                                               
## inoutgroup_effects                                           ***
## ident_c                                                      ** 
## proj_indiv_diff:stereo_indiv_diff                            *  
## proj_indiv_diff:inoutgroup_effects                           ***
## stereo_indiv_diff:inoutgroup_effects                            
## proj_indiv_diff:ident_c                                      ***
## stereo_indiv_diff:ident_c                                       
## inoutgroup_effects:ident_c                                   ***
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects         .  
## proj_indiv_diff:stereo_indiv_diff:ident_c                    ***
## proj_indiv_diff:inoutgroup_effects:ident_c                   ***
## stereo_indiv_diff:inoutgroup_effects:ident_c                    
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects:ident_c ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.206 on 7944 degrees of freedom
## Multiple R-squared:  0.4728, Adjusted R-squared:  0.4721 
## F-statistic: 647.8 on 11 and 7944 DF,  p-value: < 0.00000000000000022

linear_check <- lm(Liking ~ proj_indiv_diff*stereo_indiv_diff, mfq_individual_diff)
summary(linear_check)

## 
## Call:
## lm(formula = Liking ~ proj_indiv_diff * stereo_indiv_diff, data = mfq_individual_diff)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -4.3951 -0.9955  0.0045  1.0045  3.8465 
## 
## Coefficients: (1 not defined because of singularities)
##                                    Estimate Std. Error t value
## (Intercept)                        5.395091   0.031230 172.753
## proj_indiv_diff                   -0.961255   0.053442 -17.987
## stereo_indiv_diff                        NA         NA      NA
## proj_indiv_diff:stereo_indiv_diff  0.005448   0.015889   0.343
##                                              Pr(>|t|)    
## (Intercept)                       <0.0000000000000002 ***
## proj_indiv_diff                   <0.0000000000000002 ***
## stereo_indiv_diff                                  NA    
## proj_indiv_diff:stereo_indiv_diff               0.732    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.439 on 7953 degrees of freedom
## Multiple R-squared:  0.2486, Adjusted R-squared:  0.2484 
## F-statistic:  1315 on 2 and 7953 DF,  p-value: < 0.00000000000000022

Getting NA’s for estimates for stereotyping for individuating morals - a search online says this is because it is linearly dependent on one of the other variables (has reached singularity). After running models checking interactions between each variable and stereo_indiv_diff, it is clear for individuating morals, stereotyping and projection are linearly related (pretty much the same). This is not the case for binding morals or when the morals are combined in the overall model.

Decided not to plot individuating models, as we cannot claim this is due to counter projection, as projection is the same as stereotyping in this case.

How willing would you be to have a conversation about your political views with this person?

Overall model output

discuss_overall <- lm(WilltoDiscuss ~ proj_diff*stereo_diff*inoutgroup_effects*ident_c, downstream_analysis)
summary(discuss_overall)

## 
## Call:
## lm(formula = WilltoDiscuss ~ proj_diff * stereo_diff * inoutgroup_effects * 
##     ident_c, data = downstream_analysis)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -5.0428 -1.0559  0.2463  1.0248  3.3346 
## 
## Coefficients:
##                                                  Estimate Std. Error t value
## (Intercept)                                       5.09481    0.02811 181.226
## proj_diff                                        -0.18734    0.04651  -4.028
## stereo_diff                                      -0.33715    0.04733  -7.123
## inoutgroup_effects                                2.15681    0.05623  38.360
## ident_c                                          -0.07428    0.01513  -4.911
## proj_diff:stereo_diff                             0.17127    0.06036   2.837
## proj_diff:inoutgroup_effects                     -0.75887    0.09302  -8.159
## stereo_diff:inoutgroup_effects                   -0.62032    0.09466  -6.553
## proj_diff:ident_c                                 0.05186    0.02000   2.593
## stereo_diff:ident_c                               0.36063    0.02426  14.866
## inoutgroup_effects:ident_c                        0.22031    0.03025   7.283
## proj_diff:stereo_diff:inoutgroup_effects          0.56341    0.12072   4.667
## proj_diff:stereo_diff:ident_c                    -0.11896    0.01964  -6.057
## proj_diff:inoutgroup_effects:ident_c             -0.08602    0.04000  -2.151
## stereo_diff:inoutgroup_effects:ident_c            0.40581    0.04852   8.364
## proj_diff:stereo_diff:inoutgroup_effects:ident_c -0.14472    0.03928  -3.684
##                                                              Pr(>|t|)    
## (Intercept)                                      < 0.0000000000000002 ***
## proj_diff                                        0.000056402922172085 ***
## stereo_diff                                      0.000000000001088290 ***
## inoutgroup_effects                               < 0.0000000000000002 ***
## ident_c                                          0.000000913693464514 ***
## proj_diff:stereo_diff                                         0.00455 ** 
## proj_diff:inoutgroup_effects                     0.000000000000000359 ***
## stereo_diff:inoutgroup_effects                   0.000000000057695994 ***
## proj_diff:ident_c                                             0.00952 ** 
## stereo_diff:ident_c                              < 0.0000000000000002 ***
## inoutgroup_effects:ident_c                       0.000000000000339752 ***
## proj_diff:stereo_diff:inoutgroup_effects         0.000003076300964101 ***
## proj_diff:stereo_diff:ident_c                    0.000000001412524968 ***
## proj_diff:inoutgroup_effects:ident_c                          0.03153 *  
## stereo_diff:inoutgroup_effects:ident_c           < 0.0000000000000002 ***
## proj_diff:stereo_diff:inoutgroup_effects:ident_c              0.00023 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.679 on 19882 degrees of freedom
## Multiple R-squared:  0.2349, Adjusted R-squared:  0.2343 
## F-statistic: 406.9 on 15 and 19882 DF,  p-value: < 0.00000000000000022

Overall plots

#main effect of projection
ggplot(downstream_analysis, aes(proj_diff, WilltoDiscuss)) +
  geom_smooth(method = "lm") +
  theme_minimal() +
  labs(title = "Willingness to have a coversation with target predicted by projection of morals",
       subtitle = "Projection is the absolute value of the difference score \nbetween average MFQ scores for target and self",
       x = "Projection (Higher numbers indicate lower projection)",
       y = "Willingness")

As counter projection increased, willingness to have a conversation with the target decreased.

discuss_overall_2way_group <- effect(term = "proj_diff:inoutgroup_effects", 
                           xlevels = list(inoutgroup_effects = c(-.5,.5)), 
                           mod = discuss_overall)

discuss_overall_2way_group <- as.data.frame(discuss_overall_2way_group)

discuss_overall_2way_group$group <- as.factor(discuss_overall_2way_group$inoutgroup_effects)

ggplot(discuss_overall_2way_group, aes(proj_diff, fit, group = group, color = group)) +
 geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black", 
              aes(linetype = group)) +
  theme_minimal() +
  scale_linetype_manual("Group", breaks = c(-.5, .5), 
                       labels = c("Out-group", "In-group"),
                       values = c("solid", 
                                   "dashed")) +
  labs(title = "Willingness to have conversation with target predicted by projection and target group",
       subtitle = "Projection is the absolute value of the difference score between average \nMFQ scores for target and self. Higher scores = greater counter projection.",
       x = "Projection (Higher = more counter projection)",
       y = "Willingness")

Willingness to have a conversation increases for people in the out-group as counter projection increases, while it decreases for the in-group as counter projection occurs. One possible explanation may be that people are willing to have a conversation with their out-group who they see as morally different in order to change them, while people in their in-group who have opposite morals might be more difficult (since they seemingly chose not to have some of the morals the in-group generally seems to share).

discuss_overall_2way_ident <- effect(term = "proj_diff:ident_c", 
                           xlevels = list(ident_c = c(-1.78, 0, 1.78)), 
                           mod = discuss_overall)

discuss_overall_2way_ident <- as.data.frame(discuss_overall_2way_ident)

discuss_overall_2way_ident$Identification <- as.factor(discuss_overall_2way_ident$ident_c)

ggplot(discuss_overall_2way_ident, aes(proj_diff, fit, group = Identification, color = Identification)) +
 geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black", 
              aes(linetype = Identification)) +
  theme_minimal() +
  scale_linetype_manual("Identification", breaks = c(-1.78, 0, 1.78), 
                       labels = c("Weak (1 SD below)",
                                  "Average",
                                  "High (1 SD above)"),
                       values = c("solid", 
                                  "dashed",
                                  "dotted")) +
  labs(title = "Willingness to have conversation with target predicted by projection and identification",
       subtitle = "Projection is the absolute value of the difference score between average \nMFQ scores for target and self. Higher scores = greater counter projection.",
       x = "Projection (Higher = more counter projection)",
       y = "Willingness")

Willingness to have a conversation starts highest for all levels of identification when they project and this decreases as counter projection increases. However, as can be seen in the plot, the decrease in willingness to discuss as counter projection increases is steepest in higher identifiers, followed by average identifiers.

discuss_overall_3way <- effect(term = "proj_diff:inoutgroup_effects:ident_c", 
                           xlevels = list(ident_c = c(-1.78, 0, 1.78), 
                                         inoutgroup_effects = c(-.5,.5)), 
                           mod = discuss_overall)

discuss_overall_3way<- as.data.frame(discuss_overall_3way)

discuss_overall_3way$Group <- as.factor(discuss_overall_3way$inoutgroup_effects)
discuss_overall_3way$Identification <- as.factor(discuss_overall_3way$ident_c)

ident_label <- c("-1.78" = "Low identification \n (1 SD below)", 
                 "0" = "Average identification",
                 "1.78" = "High identificatiion \n(1 SD above)")

ggplot(discuss_overall_3way, aes(proj_diff,
                             fit, 
                             group = Group,
                             color = Group)) +
  geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black", 
              aes(linetype = Group)) +
  facet_wrap(~Identification, labeller = 
               labeller(Identification = ident_label)) +
  scale_linetype_manual("Group", breaks = c(-.5, .5), 
                       labels = c("Out-group", "In-group"),
                       values = c("solid", 
                                   "dashed")) +
  theme_minimal() +
  labs(title = "Willingness to discuss predicted by projection of morals, \nin-group identification and group membership",
       subtitle = "Controlling for stereotyping",
       x = "Projection difference scores (higher = more counter projection)",
       y = "Willingness")

Willingness to have a conversation increases the most for out-group targets as counter projection increases and in-group identification increases (weird…). In contrast, willingness to have a conversation with in-group targets decrease as both in-group identification and counter projection increase.

Binding model output

discuss_bind <- lm(WilltoDiscuss ~ proj_bind_diff*stereo_bind_diff*inoutgroup_effects*ident_c, mfq_binding_diff)
summary(discuss_bind)

## 
## Call:
## lm(formula = WilltoDiscuss ~ proj_bind_diff * stereo_bind_diff * 
##     inoutgroup_effects * ident_c, data = mfq_binding_diff)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -4.8811 -1.0349  0.2295  1.0608  3.3868 
## 
## Coefficients:
##                                                             Estimate Std. Error
## (Intercept)                                                 4.992709   0.037290
## proj_bind_diff                                             -0.194038   0.040442
## stereo_bind_diff                                            0.016543   0.042366
## inoutgroup_effects                                          2.121271   0.074579
## ident_c                                                     0.009517   0.022272
## proj_bind_diff:stereo_bind_diff                             0.003772   0.036562
## proj_bind_diff:inoutgroup_effects                          -0.241324   0.080884
## stereo_bind_diff:inoutgroup_effects                        -0.641452   0.084732
## proj_bind_diff:ident_c                                      0.059338   0.019355
## stereo_bind_diff:ident_c                                    0.119775   0.024109
## inoutgroup_effects:ident_c                                  0.027979   0.044543
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects          0.188985   0.073124
## proj_bind_diff:stereo_bind_diff:ident_c                    -0.049685   0.016036
## proj_bind_diff:inoutgroup_effects:ident_c                   0.212977   0.038711
## stereo_bind_diff:inoutgroup_effects:ident_c                 0.310620   0.048219
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects:ident_c -0.150533   0.032072
##                                                            t value
## (Intercept)                                                133.890
## proj_bind_diff                                              -4.798
## stereo_bind_diff                                             0.390
## inoutgroup_effects                                          28.443
## ident_c                                                      0.427
## proj_bind_diff:stereo_bind_diff                              0.103
## proj_bind_diff:inoutgroup_effects                           -2.984
## stereo_bind_diff:inoutgroup_effects                         -7.570
## proj_bind_diff:ident_c                                       3.066
## stereo_bind_diff:ident_c                                     4.968
## inoutgroup_effects:ident_c                                   0.628
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects           2.584
## proj_bind_diff:stereo_bind_diff:ident_c                     -3.098
## proj_bind_diff:inoutgroup_effects:ident_c                    5.502
## stereo_bind_diff:inoutgroup_effects:ident_c                  6.442
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects:ident_c  -4.694
##                                                                        Pr(>|t|)
## (Intercept)                                                < 0.0000000000000002
## proj_bind_diff                                                 0.00000162237218
## stereo_bind_diff                                                        0.69619
## inoutgroup_effects                                         < 0.0000000000000002
## ident_c                                                                 0.66914
## proj_bind_diff:stereo_bind_diff                                         0.91784
## proj_bind_diff:inoutgroup_effects                                       0.00285
## stereo_bind_diff:inoutgroup_effects                            0.00000000000004
## proj_bind_diff:ident_c                                                  0.00218
## stereo_bind_diff:ident_c                                       0.00000068586026
## inoutgroup_effects:ident_c                                              0.52992
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects                      0.00977
## proj_bind_diff:stereo_bind_diff:ident_c                                 0.00195
## proj_bind_diff:inoutgroup_effects:ident_c                      0.00000003838439
## stereo_bind_diff:inoutgroup_effects:ident_c                    0.00000000012253
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects:ident_c     0.00000271482256
##                                                               
## (Intercept)                                                ***
## proj_bind_diff                                             ***
## stereo_bind_diff                                              
## inoutgroup_effects                                         ***
## ident_c                                                       
## proj_bind_diff:stereo_bind_diff                               
## proj_bind_diff:inoutgroup_effects                          ** 
## stereo_bind_diff:inoutgroup_effects                        ***
## proj_bind_diff:ident_c                                     ** 
## stereo_bind_diff:ident_c                                   ***
## inoutgroup_effects:ident_c                                    
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects         ** 
## proj_bind_diff:stereo_bind_diff:ident_c                    ** 
## proj_bind_diff:inoutgroup_effects:ident_c                  ***
## stereo_bind_diff:inoutgroup_effects:ident_c                ***
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects:ident_c ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.685 on 11926 degrees of freedom
## Multiple R-squared:  0.2297, Adjusted R-squared:  0.2287 
## F-statistic:   237 on 15 and 11926 DF,  p-value: < 0.00000000000000022

Binding plots

#main effect of projection
ggplot(mfq_binding_diff, aes(proj_bind_diff, WilltoDiscuss)) +
  geom_smooth(method = "lm") +
  theme_minimal() +
  labs(title = "Willingness to have a coversation with target predicted by projection of binding morals",
       subtitle = "Projection is the absolute value of the difference score \nbetween average MFQ scores for target and self",
       x = "Projection (Higher numbers indicate lower projection)",
       y = "Willingness")

discuss_bind_2way_group <- effect(term = "proj_bind_diff:inoutgroup_effects", 
                           xlevels = list(inoutgroup_effects = c(-.5,.5)), 
                           mod = discuss_bind)

discuss_bind_2way_group <- as.data.frame(discuss_bind_2way_group)

discuss_bind_2way_group$group <- as.factor(discuss_bind_2way_group$inoutgroup_effects)

ggplot(discuss_bind_2way_group, aes(proj_bind_diff, fit, group = group, color = group)) +
 geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black", 
              aes(linetype = group)) +
  theme_minimal() +
  scale_linetype_manual("Group", breaks = c(-.5, .5), 
                       labels = c("Out-group", "In-group"),
                       values = c("solid", 
                                   "dashed")) +
  labs(title = "Willingness to have conversation with target predicted by projection of binding morals and target group",
       subtitle = "Projection is the absolute value of the difference score between average \nMFQ scores for target and self. Higher scores = greater counter projection.",
       x = "Projection (Higher = more counter projection)",
       y = "Willingness")

Doesn’t look like much of an interaction… but, I guess, eventually they would overlap as counter projection increases… Either way, people start out less willing to have a conversation with their out-group than their in-group. As counter projection increases, people’s willingness to discuss decreases for both in and out-group targets, but more quickly for in-group targets.

discuss_bind_2way_ident <- effect(term = "proj_bind_diff:ident_c", 
                           xlevels = list(ident_c = c(-1.78, 0, 1.78)), 
                           mod = discuss_bind)

discuss_bind_2way_ident <- as.data.frame(discuss_bind_2way_ident)

discuss_bind_2way_ident$Identification <- as.factor(discuss_bind_2way_ident$ident_c)

ggplot(discuss_bind_2way_ident, aes(proj_bind_diff, fit, group = Identification, color = Identification)) +
 geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black", 
              aes(linetype = Identification)) +
  theme_minimal() +
  scale_linetype_manual("Identification", breaks = c(-1.78, 0, 1.78), 
                       labels = c("Weak (1 SD below)",
                                  "Average",
                                  "High (1 SD above)"),
                       values = c("solid", 
                                  "dashed",
                                  "dotted")) +
  labs(title = "Willingness to have conversation with target predicted by projection of binding morals and identification",
       subtitle = "Projection is the absolute value of the difference score between average \nMFQ scores for target and self. Higher scores = greater counter projection.",
       x = "Projection (Higher = more counter projection)",
       y = "Willingness")

Similar to the last model in not being much of an interaction…

discuss_bind_3way <- effect(term = "proj_bind_diff:inoutgroup_effects:ident_c", 
                           xlevels = list(ident_c = c(-1.78, 0, 1.78), 
                                         inoutgroup_effects = c(-.5,.5)), 
                           mod = discuss_bind)

discuss_bind_3way <- as.data.frame(discuss_bind_3way)

discuss_bind_3way$Group <- as.factor(discuss_bind_3way$inoutgroup_effects)
discuss_bind_3way$Identification <- as.factor(discuss_bind_3way$ident_c)

ident_label <- c("-1.78" = "Low identification \n (1 SD below)", 
                 "0" = "Average identification",
                 "1.78" = "High identificatiion \n(1 SD above)")

ggplot(discuss_bind_3way, aes(proj_bind_diff,
                             fit, 
                             group = Group,
                             color = Group)) +
  geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black", 
              aes(linetype = Group)) +
  facet_wrap(~Identification, labeller = 
               labeller(Identification = ident_label)) +
  scale_linetype_manual("Group", breaks = c(-.5, .5), 
                       labels = c("Out-group", "In-group"),
                       values = c("solid", 
                                   "dashed")) +
  theme_minimal() +
  labs(title = "Willingness to discuss predicted by projection of binding morals, \nin-group identification and group membership",
       subtitle = "Controlling for stereotyping",
       x = "Projection difference scores (higher = more counter projection)",
       y = "Willingness")

Now this is more interesting! People who are highly identified with their in-group report the highest willingness to have a conversation with an in-group target, especially when they are projecting. Their decrease in willingness to discuss with their out-group as counter projection increases is the least steep, still resulting in a higher willingness than all other conditions. The lowest willingness to have a conversation occurred with out-group targets when participants’ highly identified with their in-group and were counter projecting. Scores for willingness to discuss for in and out-group targets were the closet to one another when participants were weakly identified with their in-group and were counter projecting.

Individuating model output

discuss_indiv <- lm(WilltoDiscuss ~ proj_indiv_diff*stereo_indiv_diff*ident_c, mfq_individual_diff)
summary(discuss_indiv)

## 
## Call:
## lm(formula = WilltoDiscuss ~ proj_indiv_diff * stereo_indiv_diff * 
##     ident_c, data = mfq_individual_diff)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -4.7927 -1.3206  0.3249  1.4753  3.8827 
## 
## Coefficients: (2 not defined because of singularities)
##                                            Estimate Std. Error t value
## (Intercept)                                5.408883   0.040091 134.915
## proj_indiv_diff                           -0.505269   0.069415  -7.279
## stereo_indiv_diff                                NA         NA      NA
## ident_c                                    0.141374   0.020390   6.934
## proj_indiv_diff:stereo_indiv_diff         -0.033770   0.020897  -1.616
## proj_indiv_diff:ident_c                   -0.024195   0.035072  -0.690
## stereo_indiv_diff:ident_c                        NA         NA      NA
## proj_indiv_diff:stereo_indiv_diff:ident_c -0.002586   0.010005  -0.258
##                                                       Pr(>|t|)    
## (Intercept)                               < 0.0000000000000002 ***
## proj_indiv_diff                              0.000000000000369 ***
## stereo_indiv_diff                                           NA    
## ident_c                                      0.000000000004424 ***
## proj_indiv_diff:stereo_indiv_diff                        0.106    
## proj_indiv_diff:ident_c                                  0.490    
## stereo_indiv_diff:ident_c                                   NA    
## proj_indiv_diff:stereo_indiv_diff:ident_c                0.796    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.828 on 7950 degrees of freedom
## Multiple R-squared:  0.09315,    Adjusted R-squared:  0.09258 
## F-statistic: 163.3 on 5 and 7950 DF,  p-value: < 0.00000000000000022

How contentious do you think a conversation about your political views would be with this person?

Overall model

content_overall <- lm(Contentious ~ proj_diff*stereo_diff*inoutgroup_effects*ident_c, downstream_analysis)
summary(content_overall)

## 
## Call:
## lm(formula = Contentious ~ proj_diff * stereo_diff * inoutgroup_effects * 
##     ident_c, data = downstream_analysis)
## 
## Residuals:
##    Min     1Q Median     3Q    Max 
## -3.930 -1.654  0.204  1.354  4.152 
## 
## Coefficients:
##                                                  Estimate Std. Error t value
## (Intercept)                                       3.75059    0.03014 124.445
## proj_diff                                         0.32046    0.04986   6.427
## stereo_diff                                       0.43910    0.05074   8.654
## inoutgroup_effects                               -1.80952    0.06028 -30.020
## ident_c                                          -0.02977    0.01622  -1.836
## proj_diff:stereo_diff                            -0.21887    0.06471  -3.382
## proj_diff:inoutgroup_effects                      0.44860    0.09972   4.499
## stereo_diff:inoutgroup_effects                    0.76520    0.10148   7.540
## proj_diff:ident_c                                 0.11142    0.02144   5.197
## stereo_diff:ident_c                               0.18932    0.02601   7.279
## inoutgroup_effects:ident_c                       -0.09929    0.03243  -3.062
## proj_diff:stereo_diff:inoutgroup_effects         -0.45101    0.12942  -3.485
## proj_diff:stereo_diff:ident_c                    -0.13300    0.02106  -6.317
## proj_diff:inoutgroup_effects:ident_c              0.22235    0.04288   5.185
## stereo_diff:inoutgroup_effects:ident_c            0.04862    0.05201   0.935
## proj_diff:stereo_diff:inoutgroup_effects:ident_c -0.08537    0.04211  -2.027
##                                                              Pr(>|t|)    
## (Intercept)                                      < 0.0000000000000002 ***
## proj_diff                                          0.0000000001328016 ***
## stereo_diff                                      < 0.0000000000000002 ***
## inoutgroup_effects                               < 0.0000000000000002 ***
## ident_c                                                      0.066426 .  
## proj_diff:stereo_diff                                        0.000720 ***
## proj_diff:inoutgroup_effects                       0.0000068773171999 ***
## stereo_diff:inoutgroup_effects                     0.0000000000000489 ***
## proj_diff:ident_c                                  0.0000002049694889 ***
## stereo_diff:ident_c                                0.0000000000003477 ***
## inoutgroup_effects:ident_c                                   0.002204 ** 
## proj_diff:stereo_diff:inoutgroup_effects                     0.000493 ***
## proj_diff:stereo_diff:ident_c                      0.0000000002729339 ***
## proj_diff:inoutgroup_effects:ident_c               0.0000002177593906 ***
## stereo_diff:inoutgroup_effects:ident_c                       0.349889    
## proj_diff:stereo_diff:inoutgroup_effects:ident_c             0.042652 *  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.8 on 19882 degrees of freedom
## Multiple R-squared:  0.1572, Adjusted R-squared:  0.1566 
## F-statistic: 247.3 on 15 and 19882 DF,  p-value: < 0.00000000000000022

Overall plots

#main effect of projection
ggplot(downstream_analysis, aes(proj_diff, Contentious)) +
  geom_smooth(method = "lm") +
  theme_minimal() +
  labs(title = "Contentiousness predicted by projection of morals",
       subtitle = "Projection is the absolute value of the difference score \nbetween average MFQ scores for target and self",
       x = "Projection (Higher numbers indicate lower projection)",
       y = "Contentousness")

As counter projection increases, predicted contentiousness with the target while have a conversation increases.

content_overall_2way_group <- effect(term = "proj_diff:inoutgroup_effects", 
                           xlevels = list(inoutgroup_effects = c(-.5,.5)), 
                           mod = content_overall)

content_overall_2way_group <- as.data.frame(content_overall_2way_group)

content_overall_2way_group$group <- as.factor(content_overall_2way_group$inoutgroup_effects)

ggplot(content_overall_2way_group, aes(proj_diff, fit, group = group, color = group)) +
 geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black", 
              aes(linetype = group)) +
  theme_minimal() +
  scale_linetype_manual("Group", breaks = c(-.5, .5), 
                       labels = c("Out-group", "In-group"),
                       values = c("solid", 
                                   "dashed")) +
  labs(title = "Contentiousness with target predicted by projection and target group",
       subtitle = "Projection is the absolute value of the difference score between average \nMFQ scores for target and self. Higher scores = greater counter projection.",
       x = "Projection (Higher = more counter projection)",
       y = "Contentiousness")

Predicted contentiousness with a target starts higher for those considering an out-group target instead of an in-group target and increases gradually as counter projection increases. However, predicted contentiousness increases more quickly as counter projection increases for those considering an in-group target.

content_overall_2way_ident <- effect(term = "proj_diff:ident_c", 
                           xlevels = list(ident_c = c(-1.78, 0, 1.78)), 
                           mod = content_overall)

content_overall_2way_ident <- as.data.frame(content_overall_2way_ident)

content_overall_2way_ident$Identification <- as.factor(content_overall_2way_ident$ident_c)

ggplot(content_overall_2way_ident, aes(proj_diff, fit, group = Identification, color = Identification)) +
 geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black", 
              aes(linetype = Identification)) +
  theme_minimal() +
  scale_linetype_manual("Identification", breaks = c(-1.78, 0, 1.78), 
                       labels = c("Weak (1 SD below)",
                                  "Average",
                                  "High (1 SD above)"),
                       values = c("solid", 
                                  "dashed",
                                  "dotted")) +
  labs(title = "Contentiousness predicted by projection and identification",
       subtitle = "Projection is the absolute value of the difference score between average \nMFQ scores for target and self. Higher scores = greater counter projection.",
       x = "Projection (Higher = more counter projection)",
       y = "Contentiousness")

Predicted contentiousness starts higher for those who are highly identified with their in-group and increases the most as counter projection increases. On the other hand, contentiousness starts the lowest for those who are weakly identified with their in-group and increases the least as counter projection increases.

content_overall_3way <- effect(term = "proj_diff:inoutgroup_effects:ident_c", 
                           xlevels = list(ident_c = c(-1.78, 0, 1.78), 
                                         inoutgroup_effects = c(-.5,.5)), 
                           mod = content_overall)

content_overall_3way <- as.data.frame(content_overall_3way)

content_overall_3way$Group <- as.factor(content_overall_3way$inoutgroup_effects)
content_overall_3way$Identification <- as.factor(content_overall_3way$ident_c)

ident_label <- c("-1.78" = "Low identification \n (1 SD below)", 
                 "0" = "Average identification",
                 "1.78" = "High identificatiion \n(1 SD above)")

ggplot(content_overall_3way, aes(proj_diff,
                             fit, 
                             group = Group,
                             color = Group)) +
  geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black", 
              aes(linetype = Group)) +
  facet_wrap(~Identification, labeller = 
               labeller(Identification = ident_label)) +
  scale_linetype_manual("Group", breaks = c(-.5, .5), 
                       labels = c("Out-group", "In-group"),
                       values = c("solid", 
                                   "dashed")) +
  theme_minimal() +
  labs(title = "Contentiousness predicted by projection of morals, \nin-group identification and group membership",
       subtitle = "Controlling for stereotyping",
       x = "Projection difference scores (higher = more counter projection)",
       y = "Contentiousness")

For those considering an out-group target, their predicted contentiousness starts higher than those considering an out-group target. As in-group identification increases, counter projection as less of an impact on predicted contentiousness with those considering out-group targets. Specifically, when weakly identified, as counter projection increases, predicted contentiousness also increases. However, when highly identified, predicted contentiousness stays the same as counter projection increases.

For those considering an in-group target, predicted contentiousness increases as counter projection increases, but more so for those who are highly identified with their in-group, followed by average identifiers, with low identifiers only showing a slight increase as counter projection increases. Interestingly, for those who are highly identified, when counter projection is high, people report higher contentiousness with the in-group target than they would with their out-group target.

Binding model

content_bind <- lm(Contentious ~ proj_bind_diff*stereo_bind_diff*ident_c*inoutgroup_effects, mfq_binding_diff)
summary(content_bind)

## 
## Call:
## lm(formula = Contentious ~ proj_bind_diff * stereo_bind_diff * 
##     ident_c * inoutgroup_effects, data = mfq_binding_diff)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -4.3664 -1.5604  0.1586  1.4965  4.0397 
## 
## Coefficients:
##                                                             Estimate Std. Error
## (Intercept)                                                 3.571486   0.039764
## proj_bind_diff                                              0.335843   0.043125
## stereo_bind_diff                                            0.402068   0.045177
## ident_c                                                     0.082495   0.023749
## inoutgroup_effects                                         -1.512059   0.079528
## proj_bind_diff:stereo_bind_diff                            -0.177484   0.038988
## proj_bind_diff:ident_c                                     -0.003608   0.020640
## stereo_bind_diff:ident_c                                    0.036241   0.025709
## proj_bind_diff:inoutgroup_effects                           0.175334   0.086250
## stereo_bind_diff:inoutgroup_effects                         0.707526   0.090354
## ident_c:inoutgroup_effects                                 -0.016837   0.047499
## proj_bind_diff:stereo_bind_diff:ident_c                    -0.008879   0.017100
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects         -0.331031   0.077976
## proj_bind_diff:ident_c:inoutgroup_effects                  -0.017748   0.041280
## stereo_bind_diff:ident_c:inoutgroup_effects                 0.015075   0.051418
## proj_bind_diff:stereo_bind_diff:ident_c:inoutgroup_effects  0.022667   0.034200
##                                                            t value
## (Intercept)                                                 89.817
## proj_bind_diff                                               7.788
## stereo_bind_diff                                             8.900
## ident_c                                                      3.474
## inoutgroup_effects                                         -19.013
## proj_bind_diff:stereo_bind_diff                             -4.552
## proj_bind_diff:ident_c                                      -0.175
## stereo_bind_diff:ident_c                                     1.410
## proj_bind_diff:inoutgroup_effects                            2.033
## stereo_bind_diff:inoutgroup_effects                          7.831
## ident_c:inoutgroup_effects                                  -0.354
## proj_bind_diff:stereo_bind_diff:ident_c                     -0.519
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects          -4.245
## proj_bind_diff:ident_c:inoutgroup_effects                   -0.430
## stereo_bind_diff:ident_c:inoutgroup_effects                  0.293
## proj_bind_diff:stereo_bind_diff:ident_c:inoutgroup_effects   0.663
##                                                                        Pr(>|t|)
## (Intercept)                                                < 0.0000000000000002
## proj_bind_diff                                              0.00000000000000739
## stereo_bind_diff                                           < 0.0000000000000002
## ident_c                                                                0.000515
## inoutgroup_effects                                         < 0.0000000000000002
## proj_bind_diff:stereo_bind_diff                             0.00000535985998455
## proj_bind_diff:ident_c                                                 0.861229
## stereo_bind_diff:ident_c                                               0.158670
## proj_bind_diff:inoutgroup_effects                                      0.042091
## stereo_bind_diff:inoutgroup_effects                         0.00000000000000527
## ident_c:inoutgroup_effects                                             0.722985
## proj_bind_diff:stereo_bind_diff:ident_c                                0.603621
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects          0.00002199662490525
## proj_bind_diff:ident_c:inoutgroup_effects                              0.667235
## stereo_bind_diff:ident_c:inoutgroup_effects                            0.769385
## proj_bind_diff:stereo_bind_diff:ident_c:inoutgroup_effects             0.507492
##                                                               
## (Intercept)                                                ***
## proj_bind_diff                                             ***
## stereo_bind_diff                                           ***
## ident_c                                                    ***
## inoutgroup_effects                                         ***
## proj_bind_diff:stereo_bind_diff                            ***
## proj_bind_diff:ident_c                                        
## stereo_bind_diff:ident_c                                      
## proj_bind_diff:inoutgroup_effects                          *  
## stereo_bind_diff:inoutgroup_effects                        ***
## ident_c:inoutgroup_effects                                    
## proj_bind_diff:stereo_bind_diff:ident_c                       
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects         ***
## proj_bind_diff:ident_c:inoutgroup_effects                     
## stereo_bind_diff:ident_c:inoutgroup_effects                   
## proj_bind_diff:stereo_bind_diff:ident_c:inoutgroup_effects    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.797 on 11926 degrees of freedom
## Multiple R-squared:  0.1601, Adjusted R-squared:  0.159 
## F-statistic: 151.5 on 15 and 11926 DF,  p-value: < 0.00000000000000022

Binding plots

ggplot(mfq_binding_diff, aes(proj_bind_diff, Contentious)) +
  geom_smooth(method = "lm") +
  theme_minimal() +
  labs(title = "Contentiousness predicted by projection of binding morals",
       subtitle = "Projection is the absolute value of the difference score \nbetween average MFQ scores for target and self",
       x = "Projection (Higher numbers indicate lower projection)",
       y = "Contentousness")

As counter projection increases, predicted contentiousness with the target while have a conversation increases.

content_bind_2way_group <- effect(term = "proj_bind_diff:inoutgroup_effects", 
                           xlevels = list(inoutgroup_effects = c(-.5,.5)), 
                           mod = content_bind)

content_bind_2way_group <- as.data.frame(content_bind_2way_group)

content_bind_2way_group$group <- as.factor(content_bind_2way_group$inoutgroup_effects)

ggplot(content_bind_2way_group, aes(proj_bind_diff, fit, group = group, color = group)) +
 geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black", 
              aes(linetype = group)) +
  theme_minimal() +
  scale_linetype_manual("Group", breaks = c(-.5, .5), 
                       labels = c("Out-group", "In-group"),
                       values = c("solid", 
                                   "dashed")) +
  labs(title = "Contentiousness with target predicted by projection of binding morals and target group",
       subtitle = "Projection is the absolute value of the difference score between average \nMFQ scores for target and self. Higher scores = greater counter projection.",
       x = "Projection (Higher = more counter projection)",
       y = "Contentiousness")

Predicted contentiousness was higher for out-group targets than in-group targets and, as counter projection of binding morals increases, contentiousness increases more for out-group targets as well.

Individuating model

content_indiv <- lm(Contentious  ~ proj_indiv_diff*stereo_indiv_diff*ident_c*inoutgroup_effects, mfq_individual_diff)
summary(content_indiv)

## 
## Call:
## lm(formula = Contentious ~ proj_indiv_diff * stereo_indiv_diff * 
##     ident_c * inoutgroup_effects, data = mfq_individual_diff)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -4.6935 -1.5171  0.1004  1.5046  4.0585 
## 
## Coefficients: (4 not defined because of singularities)
##                                                               Estimate
## (Intercept)                                                   3.748895
## proj_indiv_diff                                               0.587006
## stereo_indiv_diff                                                   NA
## ident_c                                                       0.059289
## inoutgroup_effects                                           -1.484698
## proj_indiv_diff:stereo_indiv_diff                            -0.286512
## proj_indiv_diff:ident_c                                       0.132058
## stereo_indiv_diff:ident_c                                           NA
## proj_indiv_diff:inoutgroup_effects                            1.019028
## stereo_indiv_diff:inoutgroup_effects                                NA
## ident_c:inoutgroup_effects                                    0.001465
## proj_indiv_diff:stereo_indiv_diff:ident_c                    -0.107484
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects         -0.697440
## proj_indiv_diff:ident_c:inoutgroup_effects                    0.029069
## stereo_indiv_diff:ident_c:inoutgroup_effects                        NA
## proj_indiv_diff:stereo_indiv_diff:ident_c:inoutgroup_effects -0.100359
##                                                              Std. Error t value
## (Intercept)                                                    0.044931  83.437
## proj_indiv_diff                                                0.101837   5.764
## stereo_indiv_diff                                                    NA      NA
## ident_c                                                        0.022265   2.663
## inoutgroup_effects                                             0.089862 -16.522
## proj_indiv_diff:stereo_indiv_diff                              0.054685  -5.239
## proj_indiv_diff:ident_c                                        0.051226   2.578
## stereo_indiv_diff:ident_c                                            NA      NA
## proj_indiv_diff:inoutgroup_effects                             0.203674   5.003
## stereo_indiv_diff:inoutgroup_effects                                 NA      NA
## ident_c:inoutgroup_effects                                     0.044529   0.033
## proj_indiv_diff:stereo_indiv_diff:ident_c                      0.025678  -4.186
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects           0.109369  -6.377
## proj_indiv_diff:ident_c:inoutgroup_effects                     0.102452   0.284
## stereo_indiv_diff:ident_c:inoutgroup_effects                         NA      NA
## proj_indiv_diff:stereo_indiv_diff:ident_c:inoutgroup_effects   0.051356  -1.954
##                                                                          Pr(>|t|)
## (Intercept)                                                  < 0.0000000000000002
## proj_indiv_diff                                                    0.000000008513
## stereo_indiv_diff                                                              NA
## ident_c                                                                   0.00776
## inoutgroup_effects                                           < 0.0000000000000002
## proj_indiv_diff:stereo_indiv_diff                                  0.000000165276
## proj_indiv_diff:ident_c                                                   0.00996
## stereo_indiv_diff:ident_c                                                      NA
## proj_indiv_diff:inoutgroup_effects                                 0.000000575880
## stereo_indiv_diff:inoutgroup_effects                                           NA
## ident_c:inoutgroup_effects                                                0.97375
## proj_indiv_diff:stereo_indiv_diff:ident_c                          0.000028720025
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects               0.000000000191
## proj_indiv_diff:ident_c:inoutgroup_effects                                0.77662
## stereo_indiv_diff:ident_c:inoutgroup_effects                                   NA
## proj_indiv_diff:stereo_indiv_diff:ident_c:inoutgroup_effects              0.05072
##                                                                 
## (Intercept)                                                  ***
## proj_indiv_diff                                              ***
## stereo_indiv_diff                                               
## ident_c                                                      ** 
## inoutgroup_effects                                           ***
## proj_indiv_diff:stereo_indiv_diff                            ***
## proj_indiv_diff:ident_c                                      ** 
## stereo_indiv_diff:ident_c                                       
## proj_indiv_diff:inoutgroup_effects                           ***
## stereo_indiv_diff:inoutgroup_effects                            
## ident_c:inoutgroup_effects                                      
## proj_indiv_diff:stereo_indiv_diff:ident_c                    ***
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects         ***
## proj_indiv_diff:ident_c:inoutgroup_effects                      
## stereo_indiv_diff:ident_c:inoutgroup_effects                    
## proj_indiv_diff:stereo_indiv_diff:ident_c:inoutgroup_effects .  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.799 on 7944 degrees of freedom
## Multiple R-squared:  0.1593, Adjusted R-squared:  0.1582 
## F-statistic: 136.9 on 11 and 7944 DF,  p-value: < 0.00000000000000022

How well you think the target would treat you during this conversation?

Overall model

treat_overall <- lm(Treatment ~ proj_diff*stereo_diff*inoutgroup_effects*ident_c, downstream_analysis)
summary(treat_overall)

## 
## Call:
## lm(formula = Treatment ~ proj_diff * stereo_diff * inoutgroup_effects * 
##     ident_c, data = downstream_analysis)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -4.6108 -0.8958  0.1467  0.8811  3.9993 
## 
## Coefficients:
##                                                  Estimate Std. Error t value
## (Intercept)                                       5.03368    0.02348 214.380
## proj_diff                                        -0.03487    0.03884  -0.898
## stereo_diff                                      -0.12835    0.03953  -3.247
## inoutgroup_effects                                2.06058    0.04696  43.879
## ident_c                                           0.06298    0.01263   4.985
## proj_diff:stereo_diff                            -0.18435    0.05041  -3.657
## proj_diff:inoutgroup_effects                      0.49875    0.07769   6.420
## stereo_diff:inoutgroup_effects                    0.50920    0.07906   6.441
## proj_diff:ident_c                                -0.10189    0.01670  -6.100
## stereo_diff:ident_c                               0.05234    0.02026   2.583
## inoutgroup_effects:ident_c                        0.29728    0.02527  11.766
## proj_diff:stereo_diff:inoutgroup_effects         -0.52275    0.10083  -5.185
## proj_diff:stereo_diff:ident_c                     0.01939    0.01640   1.182
## proj_diff:inoutgroup_effects:ident_c             -0.04374    0.03341  -1.309
## stereo_diff:inoutgroup_effects:ident_c           -0.08137    0.04052  -2.008
## proj_diff:stereo_diff:inoutgroup_effects:ident_c  0.05451    0.03281   1.662
##                                                              Pr(>|t|)    
## (Intercept)                                      < 0.0000000000000002 ***
## proj_diff                                                    0.369299    
## stereo_diff                                                  0.001169 ** 
## inoutgroup_effects                               < 0.0000000000000002 ***
## ident_c                                                0.000000624328 ***
## proj_diff:stereo_diff                                        0.000256 ***
## proj_diff:inoutgroup_effects                           0.000000000139 ***
## stereo_diff:inoutgroup_effects                         0.000000000122 ***
## proj_diff:ident_c                                      0.000000001083 ***
## stereo_diff:ident_c                                          0.009798 ** 
## inoutgroup_effects:ident_c                       < 0.0000000000000002 ***
## proj_diff:stereo_diff:inoutgroup_effects               0.000000218685 ***
## proj_diff:stereo_diff:ident_c                                0.237322    
## proj_diff:inoutgroup_effects:ident_c                         0.190411    
## stereo_diff:inoutgroup_effects:ident_c                       0.044643 *  
## proj_diff:stereo_diff:inoutgroup_effects:ident_c             0.096613 .  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.402 on 19882 degrees of freedom
## Multiple R-squared:  0.4673, Adjusted R-squared:  0.4669 
## F-statistic:  1163 on 15 and 19882 DF,  p-value: < 0.00000000000000022

No main effect for treatment - the only outcome in which this occurs.

Overall plots

treat_overall_2way_group <- effect(term = "proj_diff:inoutgroup_effects", 
                           xlevels = list(inoutgroup_effects = c(-.5,.5)), 
                           mod = treat_overall)

treat_overall_2way_group <- as.data.frame(treat_overall_2way_group)

treat_overall_2way_group$group <- as.factor(treat_overall_2way_group$inoutgroup_effects)

ggplot(treat_overall_2way_group, aes(proj_diff, fit, group = group, color = group)) +
 geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black", 
              aes(linetype = group)) +
  theme_minimal() +
  scale_linetype_manual("Group", breaks = c(-.5, .5), 
                       labels = c("Out-group", "In-group"),
                       values = c("solid", 
                                   "dashed")) +
  labs(title = "Treatment by target predicted by projection and target group",
       subtitle = "Projection is the absolute value of the difference score between average \nMFQ scores for target and self. Higher scores = greater counter projection.",
       x = "Projection (Higher = more counter projection)",
       y = "Treatment")

Participants expected the target to treat them better in the in-group than the out-group. As counter projection increased, expected treatment decreased for the out-group, while only barely decreasing for the in-group.

treat_overall_2way_ident <- effect(term = "proj_diff:ident_c", 
                           xlevels = list(ident_c = c(-1.78, 0, 1.78)), 
                           mod = treat_overall)

treat_overall_2way_ident <- as.data.frame(treat_overall_2way_ident)

treat_overall_2way_ident$Identification <- as.factor(treat_overall_2way_ident$ident_c)

ggplot(treat_overall_2way_ident, aes(proj_diff, fit, group = Identification, color = Identification)) +
 geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black", 
              aes(linetype = Identification)) +
  theme_minimal() +
  scale_linetype_manual("Identification", breaks = c(-1.78, 0, 1.78), 
                       labels = c("Weak (1 SD below)",
                                  "Average",
                                  "High (1 SD above)"),
                       values = c("solid", 
                                  "dashed",
                                  "dotted")) +
  labs(title = "Treatment by target predicted by projection and identification",
       subtitle = "Projection is the absolute value of the difference score between average \nMFQ scores for target and self. Higher scores = greater counter projection.",
       x = "Projection (Higher = more counter projection)",
       y = "Treatment")

For people who weakly identify with their in-group, as counter projection increases, expected treatment slightly increases. However, for people who identify with their in-group an average amount or highly, as counter projection increases, expected treatment decreases, with it decreasing the most for people who highly identified with their in-group.

Binding model

treat_bind <- lm(Treatment ~ proj_bind_diff*stereo_bind_diff*inoutgroup_effects*ident_c, mfq_binding_diff)
summary(treat_bind)

## 
## Call:
## lm(formula = Treatment ~ proj_bind_diff * stereo_bind_diff * 
##     inoutgroup_effects * ident_c, data = mfq_binding_diff)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -4.0579 -0.9026  0.0396  0.8792  4.1841 
## 
## Coefficients:
##                                                             Estimate Std. Error
## (Intercept)                                                 5.166238   0.030315
## proj_bind_diff                                             -0.052612   0.032878
## stereo_bind_diff                                           -0.102178   0.034442
## inoutgroup_effects                                          2.037037   0.060630
## ident_c                                                     0.034729   0.018106
## proj_bind_diff:stereo_bind_diff                            -0.028398   0.029724
## proj_bind_diff:inoutgroup_effects                           0.098765   0.065755
## stereo_bind_diff:inoutgroup_effects                         0.001159   0.068884
## proj_bind_diff:ident_c                                     -0.023310   0.015735
## stereo_bind_diff:ident_c                                   -0.008528   0.019600
## inoutgroup_effects:ident_c                                  0.176580   0.036212
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects          0.077894   0.059447
## proj_bind_diff:stereo_bind_diff:ident_c                     0.032445   0.013037
## proj_bind_diff:inoutgroup_effects:ident_c                   0.187069   0.031471
## stereo_bind_diff:inoutgroup_effects:ident_c                -0.050736   0.039200
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects:ident_c  0.002489   0.026073
##                                                            t value
## (Intercept)                                                170.419
## proj_bind_diff                                              -1.600
## stereo_bind_diff                                            -2.967
## inoutgroup_effects                                          33.598
## ident_c                                                      1.918
## proj_bind_diff:stereo_bind_diff                             -0.955
## proj_bind_diff:inoutgroup_effects                            1.502
## stereo_bind_diff:inoutgroup_effects                          0.017
## proj_bind_diff:ident_c                                      -1.481
## stereo_bind_diff:ident_c                                    -0.435
## inoutgroup_effects:ident_c                                   4.876
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects           1.310
## proj_bind_diff:stereo_bind_diff:ident_c                      2.489
## proj_bind_diff:inoutgroup_effects:ident_c                    5.944
## stereo_bind_diff:inoutgroup_effects:ident_c                 -1.294
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects:ident_c   0.095
##                                                                        Pr(>|t|)
## (Intercept)                                                < 0.0000000000000002
## proj_bind_diff                                                          0.10957
## stereo_bind_diff                                                        0.00302
## inoutgroup_effects                                         < 0.0000000000000002
## ident_c                                                                 0.05512
## proj_bind_diff:stereo_bind_diff                                         0.33939
## proj_bind_diff:inoutgroup_effects                                       0.13312
## stereo_bind_diff:inoutgroup_effects                                     0.98658
## proj_bind_diff:ident_c                                                  0.13852
## stereo_bind_diff:ident_c                                                0.66349
## inoutgroup_effects:ident_c                                        0.00000109495
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects                      0.19012
## proj_bind_diff:stereo_bind_diff:ident_c                                 0.01283
## proj_bind_diff:inoutgroup_effects:ident_c                         0.00000000285
## stereo_bind_diff:inoutgroup_effects:ident_c                             0.19559
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects:ident_c              0.92396
##                                                               
## (Intercept)                                                ***
## proj_bind_diff                                                
## stereo_bind_diff                                           ** 
## inoutgroup_effects                                         ***
## ident_c                                                    .  
## proj_bind_diff:stereo_bind_diff                               
## proj_bind_diff:inoutgroup_effects                             
## stereo_bind_diff:inoutgroup_effects                           
## proj_bind_diff:ident_c                                        
## stereo_bind_diff:ident_c                                      
## inoutgroup_effects:ident_c                                 ***
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects            
## proj_bind_diff:stereo_bind_diff:ident_c                    *  
## proj_bind_diff:inoutgroup_effects:ident_c                  ***
## stereo_bind_diff:inoutgroup_effects:ident_c                   
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects:ident_c    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.37 on 11926 degrees of freedom
## Multiple R-squared:  0.4915, Adjusted R-squared:  0.4909 
## F-statistic: 768.6 on 15 and 11926 DF,  p-value: < 0.00000000000000022

Binding plots

treat_bind_3way <- effect(term = "proj_bind_diff:inoutgroup_effects:ident_c", 
                           xlevels = list(ident_c = c(-1.78, 0, 1.78), 
                                         inoutgroup_effects = c(-.5,.5)), 
                           mod = treat_bind)

treat_bind_3way <- as.data.frame(treat_bind_3way)

treat_bind_3way$Group <- as.factor(treat_bind_3way$inoutgroup_effects)
treat_bind_3way$Identification <- as.factor(treat_bind_3way$ident_c)

ident_label <- c("-1.78" = "Low identification \n (1 SD below)", 
                 "0" = "Average identification",
                 "1.78" = "High identificatiion \n(1 SD above)")

ggplot(treat_bind_3way, aes(proj_bind_diff,
                             fit, 
                             group = Group,
                             color = Group)) +
  geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black", 
              aes(linetype = Group)) +
  facet_wrap(~Identification, labeller = 
               labeller(Identification = ident_label)) +
  scale_linetype_manual("Group", breaks = c(-.5, .5), 
                       labels = c("Out-group", "In-group"),
                       values = c("solid", 
                                   "dashed")) +
  theme_minimal() +
  labs(title = "Treatment by target predicted by projection of binding morals, \nin-group identification and group membership",
       subtitle = "Controlling for stereotyping",
       x = "Projection difference scores (higher = more counter projection)",
       y = "Treatment")

All treatment scores for the in-group condition were higher than those in the out-group. People in the low identification condition expected their in-group to treat them better than their out-group; However, as counter projection increased, expected treatment decreased for those in the in-group condition, while remaining the same for those in the out-group condition. For those who reported average in-group identification, as counter projection increased, people in the out-group condition reported worse treatment scores, while those in the in-group did not report a change. For those who reported high in-group identification, as count projection increased, people expected their out-group to treat them worse, while people expected their in-group to treat them better.

Individuating model

treat_indiv <- lm(Treatment  ~ proj_indiv_diff*stereo_indiv_diff*inoutgroup_effects*ident_c, mfq_individual_diff)
summary(treat_indiv)

## 
## Call:
## lm(formula = Treatment ~ proj_indiv_diff * stereo_indiv_diff * 
##     inoutgroup_effects * ident_c, data = mfq_individual_diff)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -4.3671 -0.8971  0.0353  0.8282  4.7280 
## 
## Coefficients: (4 not defined because of singularities)
##                                                               Estimate
## (Intercept)                                                   5.219881
## proj_indiv_diff                                              -0.205322
## stereo_indiv_diff                                                   NA
## inoutgroup_effects                                            1.739202
## ident_c                                                      -0.004403
## proj_indiv_diff:stereo_indiv_diff                            -0.068976
## proj_indiv_diff:inoutgroup_effects                            0.795450
## stereo_indiv_diff:inoutgroup_effects                                NA
## proj_indiv_diff:ident_c                                       0.038730
## stereo_indiv_diff:ident_c                                           NA
## inoutgroup_effects:ident_c                                    0.262681
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects         -0.143315
## proj_indiv_diff:stereo_indiv_diff:ident_c                     0.026045
## proj_indiv_diff:inoutgroup_effects:ident_c                    0.020209
## stereo_indiv_diff:inoutgroup_effects:ident_c                        NA
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects:ident_c  0.037263
##                                                              Std. Error t value
## (Intercept)                                                    0.033704 154.873
## proj_indiv_diff                                                0.076391  -2.688
## stereo_indiv_diff                                                    NA      NA
## inoutgroup_effects                                             0.067409  25.801
## ident_c                                                        0.016701  -0.264
## proj_indiv_diff:stereo_indiv_diff                              0.041021  -1.682
## proj_indiv_diff:inoutgroup_effects                             0.152783   5.206
## stereo_indiv_diff:inoutgroup_effects                                 NA      NA
## proj_indiv_diff:ident_c                                        0.038426   1.008
## stereo_indiv_diff:ident_c                                            NA      NA
## inoutgroup_effects:ident_c                                     0.033403   7.864
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects           0.082041  -1.747
## proj_indiv_diff:stereo_indiv_diff:ident_c                      0.019262   1.352
## proj_indiv_diff:inoutgroup_effects:ident_c                     0.076853   0.263
## stereo_indiv_diff:inoutgroup_effects:ident_c                         NA      NA
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects:ident_c   0.038524   0.967
##                                                                          Pr(>|t|)
## (Intercept)                                                  < 0.0000000000000002
## proj_indiv_diff                                                           0.00721
## stereo_indiv_diff                                                              NA
## inoutgroup_effects                                           < 0.0000000000000002
## ident_c                                                                   0.79207
## proj_indiv_diff:stereo_indiv_diff                                         0.09270
## proj_indiv_diff:inoutgroup_effects                            0.00000019736196861
## stereo_indiv_diff:inoutgroup_effects                                           NA
## proj_indiv_diff:ident_c                                                   0.31353
## stereo_indiv_diff:ident_c                                                      NA
## inoutgroup_effects:ident_c                                    0.00000000000000421
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects                      0.08070
## proj_indiv_diff:stereo_indiv_diff:ident_c                                 0.17636
## proj_indiv_diff:inoutgroup_effects:ident_c                                0.79259
## stereo_indiv_diff:inoutgroup_effects:ident_c                                   NA
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects:ident_c              0.33344
##                                                                 
## (Intercept)                                                  ***
## proj_indiv_diff                                              ** 
## stereo_indiv_diff                                               
## inoutgroup_effects                                           ***
## ident_c                                                         
## proj_indiv_diff:stereo_indiv_diff                            .  
## proj_indiv_diff:inoutgroup_effects                           ***
## stereo_indiv_diff:inoutgroup_effects                            
## proj_indiv_diff:ident_c                                         
## stereo_indiv_diff:ident_c                                       
## inoutgroup_effects:ident_c                                   ***
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects         .  
## proj_indiv_diff:stereo_indiv_diff:ident_c                       
## proj_indiv_diff:inoutgroup_effects:ident_c                      
## stereo_indiv_diff:inoutgroup_effects:ident_c                    
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects:ident_c    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.35 on 7944 degrees of freedom
## Multiple R-squared:  0.5074, Adjusted R-squared:  0.5067 
## F-statistic: 743.8 on 11 and 7944 DF,  p-value: < 0.00000000000000022

How valid do you think the target’s political position is?

Overall model

valid_overall <- lm(Valid ~ proj_diff*stereo_diff*inoutgroup_effects*ident_c, downstream_analysis)
summary(valid_overall)

## 
## Call:
## lm(formula = Valid ~ proj_diff * stereo_diff * inoutgroup_effects * 
##     ident_c, data = downstream_analysis)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -4.5140 -0.8753  0.2023  0.8358  4.0343 
## 
## Coefficients:
##                                                   Estimate Std. Error t value
## (Intercept)                                       5.065294   0.024030 210.791
## proj_diff                                        -0.109852   0.039753  -2.763
## stereo_diff                                      -0.262408   0.040456  -6.486
## inoutgroup_effects                                2.345308   0.048060  48.800
## ident_c                                           0.157067   0.012929  12.148
## proj_diff:stereo_diff                             0.018244   0.051594   0.354
## proj_diff:inoutgroup_effects                      0.253932   0.079506   3.194
## stereo_diff:inoutgroup_effects                    0.216820   0.080911   2.680
## proj_diff:ident_c                                -0.063391   0.017095  -3.708
## stereo_diff:ident_c                              -0.024831   0.020736  -1.197
## inoutgroup_effects:ident_c                       -0.071014   0.025858  -2.746
## proj_diff:stereo_diff:inoutgroup_effects         -0.195846   0.103189  -1.898
## proj_diff:stereo_diff:ident_c                     0.005674   0.016788   0.338
## proj_diff:inoutgroup_effects:ident_c              0.136726   0.034190   3.999
## stereo_diff:inoutgroup_effects:ident_c            0.203792   0.041472   4.914
## proj_diff:stereo_diff:inoutgroup_effects:ident_c -0.027411   0.033576  -0.816
##                                                              Pr(>|t|)    
## (Intercept)                                      < 0.0000000000000002 ***
## proj_diff                                                    0.005727 ** 
## stereo_diff                                             0.00000000009 ***
## inoutgroup_effects                               < 0.0000000000000002 ***
## ident_c                                          < 0.0000000000000002 ***
## proj_diff:stereo_diff                                        0.723638    
## proj_diff:inoutgroup_effects                                 0.001406 ** 
## stereo_diff:inoutgroup_effects                               0.007374 ** 
## proj_diff:ident_c                                            0.000209 ***
## stereo_diff:ident_c                                          0.231136    
## inoutgroup_effects:ident_c                                   0.006033 ** 
## proj_diff:stereo_diff:inoutgroup_effects                     0.057719 .  
## proj_diff:stereo_diff:ident_c                                0.735376    
## proj_diff:inoutgroup_effects:ident_c                    0.00006384085 ***
## stereo_diff:inoutgroup_effects:ident_c                  0.00000089938 ***
## proj_diff:stereo_diff:inoutgroup_effects:ident_c             0.414284    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.435 on 19882 degrees of freedom
## Multiple R-squared:  0.4734, Adjusted R-squared:  0.473 
## F-statistic:  1191 on 15 and 19882 DF,  p-value: < 0.00000000000000022

Overall plots

ggplot(downstream_analysis, aes(proj_diff, Valid)) +
  geom_smooth(method = "lm") +
  theme_minimal() +
  labs(title = "Validity of target's position predicted by projection of morals",
       subtitle = "Projection is the absolute value of the difference score \nbetween average MFQ scores for target and self",
       x = "Projection (Higher numbers indicate lower projection)",
       y = "Validity")

As counter projection increases, validity of the target’s position decreases.

valid_overall_2way_group <- effect(term = "proj_diff:inoutgroup_effects", 
                           xlevels = list(inoutgroup_effects = c(-.5,.5)), 
                           mod = valid_overall)

valid_overall_2way_group <- as.data.frame(valid_overall_2way_group)

valid_overall_2way_group$group <- as.factor(valid_overall_2way_group$inoutgroup_effects)

ggplot(valid_overall_2way_group, aes(proj_diff, fit, group = group, color = group)) +
 geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black", 
              aes(linetype = group)) +
  theme_minimal() +
  scale_linetype_manual("Group", breaks = c(-.5, .5), 
                       labels = c("Out-group", "In-group"),
                       values = c("solid", 
                                   "dashed")) +
  labs(title = "Validity of target's position predicted by projection and target group",
       subtitle = "Projection is the absolute value of the difference score between average \nMFQ scores for target and self. Higher scores = greater counter projection.",
       x = "Projection (Higher = more counter projection)",
       y = "Validity")

People report higher validity of target’s position in the in-group condition than out-group condition and, as counter projection increases, validity decreases more in the out-group condition than the in-group condition.

valid_overall_2way_ident <- effect(term = "proj_diff:ident_c", 
                           xlevels = list(ident_c = c(-1.78, 0, 1.78)), 
                           mod = valid_overall)

valid_overall_2way_ident <- as.data.frame(valid_overall_2way_ident)

valid_overall_2way_ident$Identification <- as.factor(valid_overall_2way_ident$ident_c)

ggplot(valid_overall_2way_ident, aes(proj_diff, fit, group = Identification, color = Identification)) +
 geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black", 
              aes(linetype = Identification)) +
  theme_minimal() +
  scale_linetype_manual("Identification", breaks = c(-1.78, 0, 1.78), 
                       labels = c("Weak (1 SD below)",
                                  "Average",
                                  "High (1 SD above)"),
                       values = c("solid", 
                                  "dashed",
                                  "dotted")) +
  labs(title = "Validity of target's position predicted by projection and identification",
       subtitle = "Projection is the absolute value of the difference score between average \nMFQ scores for target and self. Higher scores = greater counter projection.",
       x = "Projection (Higher = more counter projection)",
       y = "Validity")

As counter projection increases, those who report highly identifying with their in-group report the largest decrease in validity of the target’s opinion, followed by those who identify with their in-group an average amount. Those who weakly identify with their in-group actually show a slight increase in reported validity as counter projection increases.

valid_overall_3way <- effect(term = "proj_diff:inoutgroup_effects:ident_c", 
                           xlevels = list(ident_c = c(-1.78, 0, 1.78), 
                                         inoutgroup_effects = c(-.5,.5)), 
                           mod = valid_overall)

valid_overall_3way <- as.data.frame(valid_overall_3way)

valid_overall_3way$Group <- as.factor(valid_overall_3way$inoutgroup_effects)
valid_overall_3way$Identification <- as.factor(valid_overall_3way$ident_c)

ident_label <- c("-1.78" = "Low identification \n (1 SD below)", 
                 "0" = "Average identification",
                 "1.78" = "High identificatiion \n(1 SD above)")

ggplot(valid_overall_3way, aes(proj_diff,
                             fit, 
                             group = Group,
                             color = Group)) +
  geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black", 
              aes(linetype = Group)) +
  facet_wrap(~Identification, labeller = 
               labeller(Identification = ident_label)) +
  scale_linetype_manual("Group", breaks = c(-.5, .5), 
                       labels = c("Out-group", "In-group"),
                       values = c("solid", 
                                   "dashed")) +
  theme_minimal() +
  labs(title = "Validity of target's position predicted by projection of morals, \nin-group identification and group membership",
       subtitle = "Controlling for stereotyping",
       x = "Projection difference scores (higher = more counter projection)",
       y = "Validity")

For those in the in-group position, regardless of in-group identification, as counter projection increases, reported validity decreases only slightly. However, low in-group identifiers overall reported lower validity of their in-group’s opinion that high identifiers. For those in the out-group condition, as counter projection increased, low in-group identifiers report target validity increasing. However, both average and high in-group identifiers report decreases in target validity as counter projection increases, with high in-group identifiers exhibiting the largest decrease.

Binding model

valid_bind <- lm(Valid~ proj_bind_diff*stereo_bind_diff*inoutgroup_effects*ident_c, mfq_binding_diff)
summary(valid_bind)

## 
## Call:
## lm(formula = Valid ~ proj_bind_diff * stereo_bind_diff * inoutgroup_effects * 
##     ident_c, data = mfq_binding_diff)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -4.2497 -0.9416  0.0490  0.8386  4.2143 
## 
## Coefficients:
##                                                             Estimate Std. Error
## (Intercept)                                                 5.261115   0.030950
## proj_bind_diff                                             -0.217809   0.033566
## stereo_bind_diff                                           -0.096900   0.035163
## inoutgroup_effects                                          2.005359   0.061899
## ident_c                                                     0.107428   0.018485
## proj_bind_diff:stereo_bind_diff                             0.025207   0.030346
## proj_bind_diff:inoutgroup_effects                           0.301218   0.067132
## stereo_bind_diff:inoutgroup_effects                        -0.048984   0.070326
## proj_bind_diff:ident_c                                      0.002138   0.016065
## stereo_bind_diff:ident_c                                   -0.034628   0.020010
## inoutgroup_effects:ident_c                                 -0.083529   0.036970
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects          0.078145   0.060692
## proj_bind_diff:stereo_bind_diff:ident_c                     0.022681   0.013310
## proj_bind_diff:inoutgroup_effects:ident_c                   0.239194   0.032129
## stereo_bind_diff:inoutgroup_effects:ident_c                 0.095248   0.040021
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects:ident_c -0.039394   0.026619
##                                                            t value
## (Intercept)                                                169.990
## proj_bind_diff                                              -6.489
## stereo_bind_diff                                            -2.756
## inoutgroup_effects                                          32.397
## ident_c                                                      5.812
## proj_bind_diff:stereo_bind_diff                              0.831
## proj_bind_diff:inoutgroup_effects                            4.487
## stereo_bind_diff:inoutgroup_effects                         -0.697
## proj_bind_diff:ident_c                                       0.133
## stereo_bind_diff:ident_c                                    -1.730
## inoutgroup_effects:ident_c                                  -2.259
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects           1.288
## proj_bind_diff:stereo_bind_diff:ident_c                      1.704
## proj_bind_diff:inoutgroup_effects:ident_c                    7.445
## stereo_bind_diff:inoutgroup_effects:ident_c                  2.380
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects:ident_c  -1.480
##                                                                        Pr(>|t|)
## (Intercept)                                                < 0.0000000000000002
## proj_bind_diff                                                0.000000000089830
## stereo_bind_diff                                                        0.00586
## inoutgroup_effects                                         < 0.0000000000000002
## ident_c                                                       0.000000006344128
## proj_bind_diff:stereo_bind_diff                                         0.40618
## proj_bind_diff:inoutgroup_effects                             0.000007291765350
## stereo_bind_diff:inoutgroup_effects                                     0.48611
## proj_bind_diff:ident_c                                                  0.89415
## stereo_bind_diff:ident_c                                                0.08357
## inoutgroup_effects:ident_c                                              0.02388
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects                      0.19792
## proj_bind_diff:stereo_bind_diff:ident_c                                 0.08839
## proj_bind_diff:inoutgroup_effects:ident_c                     0.000000000000104
## stereo_bind_diff:inoutgroup_effects:ident_c                             0.01733
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects:ident_c              0.13892
##                                                               
## (Intercept)                                                ***
## proj_bind_diff                                             ***
## stereo_bind_diff                                           ** 
## inoutgroup_effects                                         ***
## ident_c                                                    ***
## proj_bind_diff:stereo_bind_diff                               
## proj_bind_diff:inoutgroup_effects                          ***
## stereo_bind_diff:inoutgroup_effects                           
## proj_bind_diff:ident_c                                        
## stereo_bind_diff:ident_c                                   .  
## inoutgroup_effects:ident_c                                 *  
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects            
## proj_bind_diff:stereo_bind_diff:ident_c                    .  
## proj_bind_diff:inoutgroup_effects:ident_c                  ***
## stereo_bind_diff:inoutgroup_effects:ident_c                *  
## proj_bind_diff:stereo_bind_diff:inoutgroup_effects:ident_c    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.399 on 11926 degrees of freedom
## Multiple R-squared:    0.5,  Adjusted R-squared:  0.4994 
## F-statistic: 795.1 on 15 and 11926 DF,  p-value: < 0.00000000000000022

Binding plots

#main effect of projection
ggplot(mfq_binding_diff, aes(proj_bind_diff, Valid)) +
  geom_smooth(method = "lm") +
  theme_minimal() +
  labs(title = "Validity of target's position predicted by projection of binding morals",
       subtitle = "Projection is the absolute value of the difference score \nbetween average MFQ binding scores for target and self",
       x = "Projection (Higher numbers indicate lower projection)",
       y = "Validity")

As counter projection of binding morals increases, validity decreases.

valid_bind_3way <- effect(term = "proj_bind_diff:inoutgroup_effects:ident_c", 
                           xlevels = list(ident_c = c(-1.78, 0, 1.78), 
                                         inoutgroup_effects = c(-.5,.5)), 
                           mod = valid_bind)

valid_bind_3way <- as.data.frame(valid_bind_3way)

valid_bind_3way$Group <- as.factor(valid_bind_3way$inoutgroup_effects)
valid_bind_3way$Identification <- as.factor(valid_bind_3way$ident_c)

ident_label <- c("-1.78" = "Low identification \n (1 SD below)", 
                 "0" = "Average identification",
                 "1.78" = "High identificatiion \n(1 SD above)")


#three-way interaction
ggplot(valid_bind_3way, aes(proj_bind_diff,
                             fit, 
                             group = Group,
                             color = Group)) +
  geom_smooth(method = "lm", se = FALSE, size = .7, colour = "black", 
              aes(linetype = Group)) +
  facet_wrap(~Identification, labeller = 
               labeller(Identification = ident_label)) +
  scale_linetype_manual("Group", breaks = c(-.5, .5), 
                       labels = c("Out-group", "In-group"),
                       values = c("solid", 
                                   "dashed")) +
  theme_minimal() +
  labs(title = "Validity of target's position predicted by projection of binding morals, \nin-group identification and group membership",
       subtitle = "Controlling for stereotyping",
       x = "Projection difference scores (higher = more counter projection)",
       y = "Validity")

For those who weakly identify with their in-group, people report higher validity for their in-group than their out-group, but, as counter projection increases, validity for both groups decreases by a similar amount. For people who report average identification with their in-group, they report higher validity for their in-group than their out-group and, as counter projection increases, validity for the in-group does not change, while validity for the out-group decreases. For those who highly identify with their in-group, people report higher validity for their in-group than their out-group and, as counter projection increases, validity for the in-group increases, while validity to the out-group decreases to the lowest reported validity of all conditions.

Individuating model

valid_indiv <- lm(Valid  ~ proj_indiv_diff*stereo_indiv_diff*inoutgroup_effects*ident_c, mfq_individual_diff)
summary(valid_indiv)

## 
## Call:
## lm(formula = Valid ~ proj_indiv_diff * stereo_indiv_diff * inoutgroup_effects * 
##     ident_c, data = mfq_individual_diff)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -4.1614 -0.9591  0.0237  0.8485  4.9006 
## 
## Coefficients: (4 not defined because of singularities)
##                                                              Estimate
## (Intercept)                                                   5.11705
## proj_indiv_diff                                               0.05548
## stereo_indiv_diff                                                  NA
## inoutgroup_effects                                            1.92719
## ident_c                                                       0.06032
## proj_indiv_diff:stereo_indiv_diff                            -0.22667
## proj_indiv_diff:inoutgroup_effects                            0.93074
## stereo_indiv_diff:inoutgroup_effects                               NA
## proj_indiv_diff:ident_c                                      -0.02078
## stereo_indiv_diff:ident_c                                          NA
## inoutgroup_effects:ident_c                                    0.18797
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects         -0.38056
## proj_indiv_diff:stereo_indiv_diff:ident_c                     0.06001
## proj_indiv_diff:inoutgroup_effects:ident_c                   -0.22630
## stereo_indiv_diff:inoutgroup_effects:ident_c                       NA
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects:ident_c  0.17629
##                                                              Std. Error t value
## (Intercept)                                                     0.03469 147.488
## proj_indiv_diff                                                 0.07864   0.706
## stereo_indiv_diff                                                    NA      NA
## inoutgroup_effects                                              0.06939  27.773
## ident_c                                                         0.01719   3.509
## proj_indiv_diff:stereo_indiv_diff                               0.04223  -5.368
## proj_indiv_diff:inoutgroup_effects                              0.15727   5.918
## stereo_indiv_diff:inoutgroup_effects                                 NA      NA
## proj_indiv_diff:ident_c                                         0.03956  -0.525
## stereo_indiv_diff:ident_c                                            NA      NA
## inoutgroup_effects:ident_c                                      0.03438   5.467
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects            0.08445  -4.506
## proj_indiv_diff:stereo_indiv_diff:ident_c                       0.01983   3.027
## proj_indiv_diff:inoutgroup_effects:ident_c                      0.07911  -2.861
## stereo_indiv_diff:inoutgroup_effects:ident_c                         NA      NA
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects:ident_c    0.03966   4.445
##                                                                          Pr(>|t|)
## (Intercept)                                                  < 0.0000000000000002
## proj_indiv_diff                                                          0.480472
## stereo_indiv_diff                                                              NA
## inoutgroup_effects                                           < 0.0000000000000002
## ident_c                                                                  0.000453
## proj_indiv_diff:stereo_indiv_diff                                   0.00000008185
## proj_indiv_diff:inoutgroup_effects                                  0.00000000339
## stereo_indiv_diff:inoutgroup_effects                                           NA
## proj_indiv_diff:ident_c                                                  0.599310
## stereo_indiv_diff:ident_c                                                      NA
## inoutgroup_effects:ident_c                                          0.00000004727
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects                0.00000669278
## proj_indiv_diff:stereo_indiv_diff:ident_c                                0.002481
## proj_indiv_diff:inoutgroup_effects:ident_c                               0.004240
## stereo_indiv_diff:inoutgroup_effects:ident_c                                   NA
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects:ident_c        0.00000889167
##                                                                 
## (Intercept)                                                  ***
## proj_indiv_diff                                                 
## stereo_indiv_diff                                               
## inoutgroup_effects                                           ***
## ident_c                                                      ***
## proj_indiv_diff:stereo_indiv_diff                            ***
## proj_indiv_diff:inoutgroup_effects                           ***
## stereo_indiv_diff:inoutgroup_effects                            
## proj_indiv_diff:ident_c                                         
## stereo_indiv_diff:ident_c                                       
## inoutgroup_effects:ident_c                                   ***
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects         ***
## proj_indiv_diff:stereo_indiv_diff:ident_c                    ** 
## proj_indiv_diff:inoutgroup_effects:ident_c                   ** 
## stereo_indiv_diff:inoutgroup_effects:ident_c                    
## proj_indiv_diff:stereo_indiv_diff:inoutgroup_effects:ident_c ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.389 on 7944 degrees of freedom
## Multiple R-squared:  0.507,  Adjusted R-squared:  0.5063 
## F-statistic: 742.8 on 11 and 7944 DF,  p-value: < 0.00000000000000022

Look at downstream effects with coefficients from lmer model per person

Liking

Overall

liking_coeff <- lm(Liking ~ MFQSelf_c*`MFQSelf_c:inoutgroup_effects`*`MFQSelf_c:ident_c`*`MFQSelf_c:inoutgroup_effects:ident_c`, data = downstream_coeffs)

summary(liking_coeff)

## 
## Call:
## lm(formula = Liking ~ MFQSelf_c * `MFQSelf_c:inoutgroup_effects` * 
##     `MFQSelf_c:ident_c` * `MFQSelf_c:inoutgroup_effects:ident_c`, 
##     data = downstream_coeffs)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -3.8717 -1.0608 -0.0654  1.2492  3.3006 
## 
## Coefficients: (14 not defined because of singularities)
##                                                                                                     Estimate
## (Intercept)                                                                                          4.21564
## MFQSelf_c                                                                                            2.11040
## `MFQSelf_c:inoutgroup_effects`                                                                            NA
## `MFQSelf_c:ident_c`                                                                                       NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                    NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                  NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                             NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                        NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                          NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                              NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                           NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                      NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                 NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`       NA
##                                                                                                     Std. Error
## (Intercept)                                                                                            0.07003
## MFQSelf_c                                                                                              0.24147
## `MFQSelf_c:inoutgroup_effects`                                                                              NA
## `MFQSelf_c:ident_c`                                                                                         NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                      NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                    NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                               NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                          NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                       NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                  NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                             NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                        NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`         NA
##                                                                                                     t value
## (Intercept)                                                                                           60.20
## MFQSelf_c                                                                                              8.74
## `MFQSelf_c:inoutgroup_effects`                                                                           NA
## `MFQSelf_c:ident_c`                                                                                      NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                 NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                       NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                         NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                    NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                               NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                             NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                          NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`      NA
##                                                                                                                Pr(>|t|)
## (Intercept)                                                                                         <0.0000000000000002
## MFQSelf_c                                                                                           <0.0000000000000002
## `MFQSelf_c:inoutgroup_effects`                                                                                       NA
## `MFQSelf_c:ident_c`                                                                                                  NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                               NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                             NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                                        NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                                     NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                           NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                         NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                      NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                 NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                            NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                  NA
##                                                                                                        
## (Intercept)                                                                                         ***
## MFQSelf_c                                                                                           ***
## `MFQSelf_c:inoutgroup_effects`                                                                         
## `MFQSelf_c:ident_c`                                                                                    
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                 
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                               
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                          
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                     
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                       
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                  
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                             
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                           
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                        
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                   
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`              
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.576 on 667 degrees of freedom
##   (5 observations deleted due to missingness)
## Multiple R-squared:  0.1028, Adjusted R-squared:  0.1014 
## F-statistic: 76.39 on 1 and 667 DF,  p-value: < 0.00000000000000022

ggplot(downstream_coeffs,aes(MFQSelf_c, Liking)) +
  geom_smooth(method = "lm") +
  labs(title = "Liking predicted by projection coefficient",
       subtitle = "Coefficient from initial study model controlling for stereotyping",
       x = "Projection coefficient") +
  theme_minimal()

Liking including similiarity difference scores

diffs <- downstream_analysis %>% 
  select(subid, proj_diff, stereo_diff)

coeffs_and_diffs <- left_join(downstream_coeffs, diffs) %>%  unique()

liking_codiff <- lm(Liking ~ MFQSelf_c*proj_diff, coeffs_and_diffs)
summary(liking_codiff)

## 
## Call:
## lm(formula = Liking ~ MFQSelf_c * proj_diff, data = coeffs_and_diffs)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -4.0886 -1.0369  0.0241  1.1853  3.9471 
## 
## Coefficients:
##                     Estimate Std. Error t value             Pr(>|t|)    
## (Intercept)           4.6136     0.1144  40.334 < 0.0000000000000002 ***
## MFQSelf_c             1.8815     0.3948   4.765           0.00000232 ***
## proj_diff            -0.6135     0.1292  -4.748           0.00000252 ***
## MFQSelf_c:proj_diff  -0.1288     0.5037  -0.256                0.798    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.546 on 665 degrees of freedom
##   (5 observations deleted due to missingness)
## Multiple R-squared:  0.1397, Adjusted R-squared:  0.1358 
## F-statistic: 35.99 on 3 and 665 DF,  p-value: < 0.00000000000000022

Two main effects but no interaction for liking. By the negative coefficient for the similarity score, we can tell that as dissimilarity increases, liking decreases (like the plot we did for the downstream effects using difference scores before).

Liking with binding model coefficients

#getting the coefficients from initial study with stereotyping
coeffs_bind <- coef(is_mfqbinding)$subid

subid <- rownames(coeffs_bind)

coeffs_bind <- cbind(id = subid, coeffs_bind)

coeffs_bind <- coeffs_bind %>% mutate(subid = as.numeric(id))

downstream_coeffs_bind <- left_join(coeffs_bind, tomerge) %>% select(-id) %>% unique()

liking_coeff_bind <- lm(Liking ~ MFQSelf_c*`MFQSelf_c:inoutgroup_effects`*`MFQSelf_c:ident_c`*`MFQSelf_c:inoutgroup_effects:ident_c`, data = downstream_coeffs_bind)
summary(liking_coeff_bind)

## 
## Call:
## lm(formula = Liking ~ MFQSelf_c * `MFQSelf_c:inoutgroup_effects` * 
##     `MFQSelf_c:ident_c` * `MFQSelf_c:inoutgroup_effects:ident_c`, 
##     data = downstream_coeffs_bind)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -4.0808 -0.9415 -0.1127  1.3590  3.2279 
## 
## Coefficients: (14 not defined because of singularities)
##                                                                                                     Estimate
## (Intercept)                                                                                          4.23230
## MFQSelf_c                                                                                            2.26496
## `MFQSelf_c:inoutgroup_effects`                                                                            NA
## `MFQSelf_c:ident_c`                                                                                       NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                    NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                  NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                             NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                        NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                          NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                              NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                           NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                      NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                 NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`       NA
##                                                                                                     Std. Error
## (Intercept)                                                                                            0.08132
## MFQSelf_c                                                                                              0.40960
## `MFQSelf_c:inoutgroup_effects`                                                                              NA
## `MFQSelf_c:ident_c`                                                                                         NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                      NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                    NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                               NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                          NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                       NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                  NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                             NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                        NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`         NA
##                                                                                                     t value
## (Intercept)                                                                                           52.05
## MFQSelf_c                                                                                              5.53
## `MFQSelf_c:inoutgroup_effects`                                                                           NA
## `MFQSelf_c:ident_c`                                                                                      NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                 NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                       NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                         NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                    NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                               NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                             NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                          NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`      NA
##                                                                                                                 Pr(>|t|)
## (Intercept)                                                                                         < 0.0000000000000002
## MFQSelf_c                                                                                                   0.0000000461
## `MFQSelf_c:inoutgroup_effects`                                                                                        NA
## `MFQSelf_c:ident_c`                                                                                                   NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                              NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                                         NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                                    NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                                      NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                 NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                            NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                          NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                       NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                  NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                             NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                   NA
##                                                                                                        
## (Intercept)                                                                                         ***
## MFQSelf_c                                                                                           ***
## `MFQSelf_c:inoutgroup_effects`                                                                         
## `MFQSelf_c:ident_c`                                                                                    
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                 
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                               
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                          
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                     
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                       
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                  
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                             
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                           
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                        
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                   
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`              
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.627 on 667 degrees of freedom
## Multiple R-squared:  0.04383,    Adjusted R-squared:  0.0424 
## F-statistic: 30.58 on 1 and 667 DF,  p-value: 0.00000004608

ggplot(downstream_coeffs_bind,aes(MFQSelf_c, Liking)) +
  geom_smooth(method = "lm") +
  labs(title = "Liking predicted by projection coefficient",
       subtitle = "Coefficient from initial study with binding data only controlling for stereotyping",
       x = "Projection coefficient") +
  theme_minimal()

This looks pretty much the same as the overall model

Liking with individuating model coefficients

liking_coeff_indiv <- lm(Liking ~ MFQSelf_c*`MFQSelf_c:inoutgroup_effects`*`MFQSelf_c:ident_c`*`MFQSelf_c:inoutgroup_effects:ident_c`, data = downstream_coeffs_indiv)
summary(liking_coeff_indiv)

## 
## Call:
## lm(formula = Liking ~ MFQSelf_c * `MFQSelf_c:inoutgroup_effects` * 
##     `MFQSelf_c:ident_c` * `MFQSelf_c:inoutgroup_effects:ident_c`, 
##     data = downstream_coeffs_indiv)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -4.0738 -0.8740 -0.1002  1.3210  3.1777 
## 
## Coefficients: (14 not defined because of singularities)
##                                                                                                     Estimate
## (Intercept)                                                                                          4.10544
## MFQSelf_c                                                                                            1.94964
## `MFQSelf_c:inoutgroup_effects`                                                                            NA
## `MFQSelf_c:ident_c`                                                                                       NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                    NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                  NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                             NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                        NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                          NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                              NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                           NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                      NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                 NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`       NA
##                                                                                                     Std. Error
## (Intercept)                                                                                            0.09869
## MFQSelf_c                                                                                              0.35985
## `MFQSelf_c:inoutgroup_effects`                                                                              NA
## `MFQSelf_c:ident_c`                                                                                         NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                      NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                    NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                               NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                          NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                       NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                  NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                             NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                        NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`         NA
##                                                                                                     t value
## (Intercept)                                                                                          41.598
## MFQSelf_c                                                                                             5.418
## `MFQSelf_c:inoutgroup_effects`                                                                           NA
## `MFQSelf_c:ident_c`                                                                                      NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                 NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                       NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                         NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                    NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                               NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                             NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                          NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`      NA
##                                                                                                                 Pr(>|t|)
## (Intercept)                                                                                         < 0.0000000000000002
## MFQSelf_c                                                                                                   0.0000000843
## `MFQSelf_c:inoutgroup_effects`                                                                                        NA
## `MFQSelf_c:ident_c`                                                                                                   NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                              NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                                         NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                                    NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                                      NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                 NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                            NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                          NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                       NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                  NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                             NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                   NA
##                                                                                                        
## (Intercept)                                                                                         ***
## MFQSelf_c                                                                                           ***
## `MFQSelf_c:inoutgroup_effects`                                                                         
## `MFQSelf_c:ident_c`                                                                                    
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                 
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                               
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                          
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                     
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                       
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                  
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                             
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                           
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                        
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                   
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`              
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.629 on 667 degrees of freedom
## Multiple R-squared:  0.04215,    Adjusted R-squared:  0.04072 
## F-statistic: 29.35 on 1 and 667 DF,  p-value: 0.0000000843

ggplot(downstream_coeffs_indiv,aes(MFQSelf_c, Liking)) +
  geom_smooth(method = "lm") +
  labs(title = "Liking predicted by projection coefficient",
       subtitle = "Coefficient from initial study with individuating data only controlling for stereotyping",
       x = "Projection coefficient") +
  theme_minimal()

Also pretty much the same as the overall model

Willingness to have discussion

discuss_coeffs <- lm(WilltoDiscuss ~ MFQSelf_c*`MFQSelf_c:inoutgroup_effects`*`MFQSelf_c:ident_c`*`MFQSelf_c:inoutgroup_effects:ident_c`, data = downstream_coeffs)
summary(discuss_coeffs)

## 
## Call:
## lm(formula = WilltoDiscuss ~ MFQSelf_c * `MFQSelf_c:inoutgroup_effects` * 
##     `MFQSelf_c:ident_c` * `MFQSelf_c:inoutgroup_effects:ident_c`, 
##     data = downstream_coeffs)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -4.5085 -1.1812  0.2831  1.7153  2.6674 
## 
## Coefficients: (14 not defined because of singularities)
##                                                                                                     Estimate
## (Intercept)                                                                                          4.75088
## MFQSelf_c                                                                                            0.93792
## `MFQSelf_c:inoutgroup_effects`                                                                            NA
## `MFQSelf_c:ident_c`                                                                                       NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                    NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                  NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                             NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                        NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                          NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                              NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                           NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                      NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                 NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`       NA
##                                                                                                     Std. Error
## (Intercept)                                                                                            0.08478
## MFQSelf_c                                                                                              0.29231
## `MFQSelf_c:inoutgroup_effects`                                                                              NA
## `MFQSelf_c:ident_c`                                                                                         NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                      NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                    NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                               NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                          NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                       NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                  NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                             NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                        NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`         NA
##                                                                                                     t value
## (Intercept)                                                                                          56.039
## MFQSelf_c                                                                                             3.209
## `MFQSelf_c:inoutgroup_effects`                                                                           NA
## `MFQSelf_c:ident_c`                                                                                      NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                 NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                       NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                         NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                    NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                               NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                             NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                          NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`      NA
##                                                                                                                Pr(>|t|)
## (Intercept)                                                                                         <0.0000000000000002
## MFQSelf_c                                                                                                        0.0014
## `MFQSelf_c:inoutgroup_effects`                                                                                       NA
## `MFQSelf_c:ident_c`                                                                                                  NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                               NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                             NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                                        NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                                     NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                           NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                         NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                      NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                 NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                            NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                  NA
##                                                                                                        
## (Intercept)                                                                                         ***
## MFQSelf_c                                                                                           ** 
## `MFQSelf_c:inoutgroup_effects`                                                                         
## `MFQSelf_c:ident_c`                                                                                    
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                 
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                               
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                          
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                     
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                       
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                  
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                             
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                           
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                        
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                   
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`              
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.908 on 667 degrees of freedom
##   (5 observations deleted due to missingness)
## Multiple R-squared:  0.0152, Adjusted R-squared:  0.01372 
## F-statistic:  10.3 on 1 and 667 DF,  p-value: 0.001398

ggplot(downstream_coeffs,aes(MFQSelf_c, WilltoDiscuss)) +
  geom_smooth(method = "lm") +
  labs(title = "Willigness to have a conversation predicted by projection coefficient",
       subtitle = "Coefficient from initial study controlling for stereotyping",
       x = "Projection coefficient",
       y = "Willingness") +
  theme_minimal()

Willingness with difference scores too

discuss_codiff <- lm(WilltoDiscuss ~ MFQSelf_c*proj_diff, coeffs_and_diffs)
summary(discuss_codiff)

## 
## Call:
## lm(formula = WilltoDiscuss ~ MFQSelf_c * proj_diff, data = coeffs_and_diffs)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -4.7063 -1.2055  0.3149  1.5950  3.4968 
## 
## Coefficients:
##                     Estimate Std. Error t value             Pr(>|t|)    
## (Intercept)           5.0611     0.1395  36.281 < 0.0000000000000002 ***
## MFQSelf_c             1.1180     0.4815   2.322              0.02055 *  
## proj_diff            -0.5069     0.1576  -3.217              0.00136 ** 
## MFQSelf_c:proj_diff  -0.7608     0.6143  -1.238              0.21601    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.885 on 665 degrees of freedom
##   (5 observations deleted due to missingness)
## Multiple R-squared:  0.04166,    Adjusted R-squared:  0.03734 
## F-statistic: 9.637 on 3 and 665 DF,  p-value: 0.00000311

Still significant after including the similarity scores, which, again, are also significant but the interaction between the two is not.

Willingness with binding data

discuss_coeff_bind <- lm(WilltoDiscuss ~ MFQSelf_c*`MFQSelf_c:inoutgroup_effects`*`MFQSelf_c:ident_c`*`MFQSelf_c:inoutgroup_effects:ident_c`, data = downstream_coeffs_bind)
summary(discuss_coeff_bind)

## 
## Call:
## lm(formula = WilltoDiscuss ~ MFQSelf_c * `MFQSelf_c:inoutgroup_effects` * 
##     `MFQSelf_c:ident_c` * `MFQSelf_c:inoutgroup_effects:ident_c`, 
##     data = downstream_coeffs_bind)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -4.3181 -1.0979  0.2413  1.7868  2.5792 
## 
## Coefficients: (14 not defined because of singularities)
##                                                                                                     Estimate
## (Intercept)                                                                                          4.75351
## MFQSelf_c                                                                                            1.04463
## `MFQSelf_c:inoutgroup_effects`                                                                            NA
## `MFQSelf_c:ident_c`                                                                                       NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                    NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                  NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                             NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                        NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                          NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                              NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                           NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                      NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                 NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`       NA
##                                                                                                     Std. Error
## (Intercept)                                                                                            0.09575
## MFQSelf_c                                                                                              0.48233
## `MFQSelf_c:inoutgroup_effects`                                                                              NA
## `MFQSelf_c:ident_c`                                                                                         NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                      NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                    NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                               NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                          NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                       NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                  NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                             NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                        NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`         NA
##                                                                                                     t value
## (Intercept)                                                                                          49.643
## MFQSelf_c                                                                                             2.166
## `MFQSelf_c:inoutgroup_effects`                                                                           NA
## `MFQSelf_c:ident_c`                                                                                      NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                 NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                       NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                         NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                    NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                               NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                             NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                          NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`      NA
##                                                                                                                Pr(>|t|)
## (Intercept)                                                                                         <0.0000000000000002
## MFQSelf_c                                                                                                        0.0307
## `MFQSelf_c:inoutgroup_effects`                                                                                       NA
## `MFQSelf_c:ident_c`                                                                                                  NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                               NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                             NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                                        NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                                     NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                           NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                         NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                      NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                 NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                            NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                  NA
##                                                                                                        
## (Intercept)                                                                                         ***
## MFQSelf_c                                                                                           *  
## `MFQSelf_c:inoutgroup_effects`                                                                         
## `MFQSelf_c:ident_c`                                                                                    
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                 
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                               
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                          
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                     
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                       
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                  
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                             
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                           
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                        
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                   
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`              
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.916 on 667 degrees of freedom
## Multiple R-squared:  0.006984,   Adjusted R-squared:  0.005495 
## F-statistic: 4.691 on 1 and 667 DF,  p-value: 0.03068

ggplot(downstream_coeffs_bind,aes(MFQSelf_c, WilltoDiscuss)) +
  geom_smooth(method = "lm") +
  labs(title = "Willingness to have discussion predicted by projection coefficient",
       subtitle = "Coefficient from initial study with binding data only controlling for stereotyping",
       x = "Projection coefficient",
       y = "Willingness") +
  theme_minimal()

Willingness with individuating data

discuss_coeff_indiv <- lm(WilltoDiscuss ~ MFQSelf_c*`MFQSelf_c:inoutgroup_effects`*`MFQSelf_c:ident_c`*`MFQSelf_c:inoutgroup_effects:ident_c`, data = downstream_coeffs_indiv)
summary(discuss_coeff_indiv)

## 
## Call:
## lm(formula = WilltoDiscuss ~ MFQSelf_c * `MFQSelf_c:inoutgroup_effects` * 
##     `MFQSelf_c:ident_c` * `MFQSelf_c:inoutgroup_effects:ident_c`, 
##     data = downstream_coeffs_indiv)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -4.3569 -1.0171  0.1831  1.8784  2.5008 
## 
## Coefficients: (14 not defined because of singularities)
##                                                                                                     Estimate
## (Intercept)                                                                                           4.7324
## MFQSelf_c                                                                                             0.7222
## `MFQSelf_c:inoutgroup_effects`                                                                            NA
## `MFQSelf_c:ident_c`                                                                                       NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                    NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                  NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                             NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                        NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                          NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                              NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                           NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                      NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                 NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`       NA
##                                                                                                     Std. Error
## (Intercept)                                                                                             0.1163
## MFQSelf_c                                                                                               0.4239
## `MFQSelf_c:inoutgroup_effects`                                                                              NA
## `MFQSelf_c:ident_c`                                                                                         NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                      NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                    NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                               NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                          NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                       NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                  NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                             NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                        NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`         NA
##                                                                                                     t value
## (Intercept)                                                                                          40.702
## MFQSelf_c                                                                                             1.703
## `MFQSelf_c:inoutgroup_effects`                                                                           NA
## `MFQSelf_c:ident_c`                                                                                      NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                 NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                       NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                         NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                    NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                               NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                             NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                          NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`      NA
##                                                                                                                Pr(>|t|)
## (Intercept)                                                                                         <0.0000000000000002
## MFQSelf_c                                                                                                        0.0889
## `MFQSelf_c:inoutgroup_effects`                                                                                       NA
## `MFQSelf_c:ident_c`                                                                                                  NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                               NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                             NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                                        NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                                     NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                           NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                         NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                      NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                 NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                            NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                  NA
##                                                                                                        
## (Intercept)                                                                                         ***
## MFQSelf_c                                                                                           .  
## `MFQSelf_c:inoutgroup_effects`                                                                         
## `MFQSelf_c:ident_c`                                                                                    
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                 
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                               
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                          
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                     
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                       
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                  
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                             
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                           
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                        
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                   
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`              
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.919 on 667 degrees of freedom
## Multiple R-squared:  0.004332,   Adjusted R-squared:  0.002839 
## F-statistic: 2.902 on 1 and 667 DF,  p-value: 0.08895

ggplot(downstream_coeffs_indiv,aes(MFQSelf_c, WilltoDiscuss)) +
  geom_smooth(method = "lm") +
  labs(title = "Willingness to have discussion predicted by projection coefficient",
       subtitle = "Coefficient from initial study with individuating data only controlling for stereotyping",
       x = "Projection coefficient",
       y = "Willingness") +
  theme_minimal()

The effect was there with the binding data, but is only marginal with the individuating data.

Contentiousness predicted by coefficients

content_coeffs <- lm(Contentious ~ MFQSelf_c*`MFQSelf_c:inoutgroup_effects`*`MFQSelf_c:ident_c`*`MFQSelf_c:inoutgroup_effects:ident_c`, data = downstream_coeffs)
summary(content_coeffs)

## 
## Call:
## lm(formula = Contentious ~ MFQSelf_c * `MFQSelf_c:inoutgroup_effects` * 
##     `MFQSelf_c:ident_c` * `MFQSelf_c:inoutgroup_effects:ident_c`, 
##     data = downstream_coeffs)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -4.3225 -1.6000  0.0448  1.6434  4.1309 
## 
## Coefficients: (14 not defined because of singularities)
##                                                                                                     Estimate
## (Intercept)                                                                                          4.30714
## MFQSelf_c                                                                                           -1.95072
## `MFQSelf_c:inoutgroup_effects`                                                                            NA
## `MFQSelf_c:ident_c`                                                                                       NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                    NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                  NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                             NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                        NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                          NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                              NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                           NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                      NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                 NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`       NA
##                                                                                                     Std. Error
## (Intercept)                                                                                            0.08439
## MFQSelf_c                                                                                              0.29096
## `MFQSelf_c:inoutgroup_effects`                                                                              NA
## `MFQSelf_c:ident_c`                                                                                         NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                      NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                    NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                               NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                          NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                       NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                  NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                             NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                        NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`         NA
##                                                                                                     t value
## (Intercept)                                                                                          51.041
## MFQSelf_c                                                                                            -6.704
## `MFQSelf_c:inoutgroup_effects`                                                                           NA
## `MFQSelf_c:ident_c`                                                                                      NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                 NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                       NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                         NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                    NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                               NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                             NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                          NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`      NA
##                                                                                                                 Pr(>|t|)
## (Intercept)                                                                                         < 0.0000000000000002
## MFQSelf_c                                                                                                0.0000000000431
## `MFQSelf_c:inoutgroup_effects`                                                                                        NA
## `MFQSelf_c:ident_c`                                                                                                   NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                              NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                                         NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                                    NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                                      NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                 NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                            NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                          NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                       NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                  NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                             NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                   NA
##                                                                                                        
## (Intercept)                                                                                         ***
## MFQSelf_c                                                                                           ***
## `MFQSelf_c:inoutgroup_effects`                                                                         
## `MFQSelf_c:ident_c`                                                                                    
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                 
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                               
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                          
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                     
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                       
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                  
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                             
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                           
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                        
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                   
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`              
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.899 on 667 degrees of freedom
##   (5 observations deleted due to missingness)
## Multiple R-squared:  0.06313,    Adjusted R-squared:  0.06173 
## F-statistic: 44.95 on 1 and 667 DF,  p-value: 0.00000000004313

ggplot(downstream_coeffs,aes(MFQSelf_c, Contentious)) +
  geom_smooth(method = "lm") +
  labs(title = "Contentiousness predicted by projection coefficient",
       subtitle = "Coefficient from initial study controlling for stereotyping",
       x = "Projection coefficient",
       y = "Contentiousness") +
  theme_minimal()

Contentiousness with difference scores too

content_codiff <- lm(Contentious ~ MFQSelf_c*proj_diff, coeffs_and_diffs)
summary(content_codiff)

## 
## Call:
## lm(formula = Contentious ~ MFQSelf_c * proj_diff, data = coeffs_and_diffs)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -4.4950 -1.6050  0.1114  1.5978  3.8107 
## 
## Coefficients:
##                     Estimate Std. Error t value             Pr(>|t|)    
## (Intercept)          4.10228    0.14030  29.240 < 0.0000000000000002 ***
## MFQSelf_c           -1.75318    0.48427  -3.620             0.000317 ***
## proj_diff            0.30945    0.15850   1.952             0.051307 .  
## MFQSelf_c:proj_diff -0.08051    0.61784  -0.130             0.896364    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.896 on 665 degrees of freedom
##   (5 observations deleted due to missingness)
## Multiple R-squared:  0.06925,    Adjusted R-squared:  0.06506 
## F-statistic: 16.49 on 3 and 665 DF,  p-value: 0.0000000002392

Projection is significant, but the difference score is only marginal.

Contentiousness with binding data

content_coeff_bind <- lm(Contentious ~ MFQSelf_c*`MFQSelf_c:inoutgroup_effects`*`MFQSelf_c:ident_c`*`MFQSelf_c:inoutgroup_effects:ident_c`, data = downstream_coeffs_bind)
summary(content_coeff_bind)

## 
## Call:
## lm(formula = Contentious ~ MFQSelf_c * `MFQSelf_c:inoutgroup_effects` * 
##     `MFQSelf_c:ident_c` * `MFQSelf_c:inoutgroup_effects:ident_c`, 
##     data = downstream_coeffs_bind)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -3.8278 -1.6903  0.0235  1.7339  3.5124 
## 
## Coefficients: (14 not defined because of singularities)
##                                                                                                     Estimate
## (Intercept)                                                                                          4.30173
## MFQSelf_c                                                                                           -2.17300
## `MFQSelf_c:inoutgroup_effects`                                                                            NA
## `MFQSelf_c:ident_c`                                                                                       NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                    NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                  NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                             NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                        NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                          NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                              NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                           NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                      NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                 NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`       NA
##                                                                                                     Std. Error
## (Intercept)                                                                                            0.09663
## MFQSelf_c                                                                                              0.48674
## `MFQSelf_c:inoutgroup_effects`                                                                              NA
## `MFQSelf_c:ident_c`                                                                                         NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                      NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                    NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                               NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                          NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                       NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                  NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                             NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                        NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`         NA
##                                                                                                     t value
## (Intercept)                                                                                          44.518
## MFQSelf_c                                                                                            -4.464
## `MFQSelf_c:inoutgroup_effects`                                                                           NA
## `MFQSelf_c:ident_c`                                                                                      NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                 NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                       NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                         NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                    NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                               NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                             NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                          NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`      NA
##                                                                                                                 Pr(>|t|)
## (Intercept)                                                                                         < 0.0000000000000002
## MFQSelf_c                                                                                                     0.00000943
## `MFQSelf_c:inoutgroup_effects`                                                                                        NA
## `MFQSelf_c:ident_c`                                                                                                   NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                              NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                                         NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                                    NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                                      NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                 NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                            NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                          NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                       NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                  NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                             NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                   NA
##                                                                                                        
## (Intercept)                                                                                         ***
## MFQSelf_c                                                                                           ***
## `MFQSelf_c:inoutgroup_effects`                                                                         
## `MFQSelf_c:ident_c`                                                                                    
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                 
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                               
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                          
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                     
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                       
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                  
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                             
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                           
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                        
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                   
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`              
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.934 on 667 degrees of freedom
## Multiple R-squared:  0.02901,    Adjusted R-squared:  0.02756 
## F-statistic: 19.93 on 1 and 667 DF,  p-value: 0.000009426

ggplot(downstream_coeffs_bind,aes(MFQSelf_c, Contentious)) +
  geom_smooth(method = "lm") +
  labs(title = "Contentiousness predicted by projection coefficient",
       subtitle = "Coefficient from initial study with binding data only controlling for stereotyping",
       x = "Projection coefficient",
       y = "Contentiousness") +
  theme_minimal()

Contentiousness with individuating data

content_coeff_indiv <- lm(Contentious ~ MFQSelf_c*`MFQSelf_c:inoutgroup_effects`*`MFQSelf_c:ident_c`*`MFQSelf_c:inoutgroup_effects:ident_c`, data = downstream_coeffs_indiv)
summary(content_coeff_indiv)

## 
## Call:
## lm(formula = Contentious ~ MFQSelf_c * `MFQSelf_c:inoutgroup_effects` * 
##     `MFQSelf_c:ident_c` * `MFQSelf_c:inoutgroup_effects:ident_c`, 
##     data = downstream_coeffs_indiv)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -3.8421 -1.7297  0.0059  1.7249  4.0858 
## 
## Coefficients: (14 not defined because of singularities)
##                                                                                                     Estimate
## (Intercept)                                                                                           4.3884
## MFQSelf_c                                                                                            -1.7047
## `MFQSelf_c:inoutgroup_effects`                                                                            NA
## `MFQSelf_c:ident_c`                                                                                       NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                    NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                  NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                             NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                        NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                          NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                              NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                           NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                      NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                 NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`       NA
##                                                                                                     Std. Error
## (Intercept)                                                                                             0.1175
## MFQSelf_c                                                                                               0.4285
## `MFQSelf_c:inoutgroup_effects`                                                                              NA
## `MFQSelf_c:ident_c`                                                                                         NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                      NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                    NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                               NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                          NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                       NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                  NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                             NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                        NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`         NA
##                                                                                                     t value
## (Intercept)                                                                                          37.339
## MFQSelf_c                                                                                            -3.978
## `MFQSelf_c:inoutgroup_effects`                                                                           NA
## `MFQSelf_c:ident_c`                                                                                      NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                 NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                       NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                         NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                    NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                               NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                             NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                          NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`      NA
##                                                                                                                 Pr(>|t|)
## (Intercept)                                                                                         < 0.0000000000000002
## MFQSelf_c                                                                                                      0.0000771
## `MFQSelf_c:inoutgroup_effects`                                                                                        NA
## `MFQSelf_c:ident_c`                                                                                                   NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                              NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                                         NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                                    NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                                      NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                 NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                            NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                          NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                       NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                  NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                             NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                   NA
##                                                                                                        
## (Intercept)                                                                                         ***
## MFQSelf_c                                                                                           ***
## `MFQSelf_c:inoutgroup_effects`                                                                         
## `MFQSelf_c:ident_c`                                                                                    
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                 
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                               
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                          
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                     
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                       
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                  
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                             
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                           
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                        
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                   
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`              
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.939 on 667 degrees of freedom
## Multiple R-squared:  0.02317,    Adjusted R-squared:  0.02171 
## F-statistic: 15.82 on 1 and 667 DF,  p-value: 0.00007712

ggplot(downstream_coeffs_indiv,aes(MFQSelf_c, Contentious)) +
  geom_smooth(method = "lm") +
  labs(title = "Contentiousness predicted by projection coefficient",
       subtitle = "Coefficient from initial study with individuating data only controlling for stereotyping",
       x = "Projection coefficient",
       y = "Contentiousness") +
  theme_minimal()

The models for individuating and binding data don’t differ much from the overall model.

Treatment predicted by coefficients

treat_coeffs <- lm(Treatment ~ MFQSelf_c*`MFQSelf_c:inoutgroup_effects`*`MFQSelf_c:ident_c`*`MFQSelf_c:inoutgroup_effects:ident_c`, data = downstream_coeffs)
summary(treat_coeffs)

## 
## Call:
## lm(formula = Treatment ~ MFQSelf_c * `MFQSelf_c:inoutgroup_effects` * 
##     `MFQSelf_c:ident_c` * `MFQSelf_c:inoutgroup_effects:ident_c`, 
##     data = downstream_coeffs)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -5.1786 -1.4518  0.4243  1.5159  3.6319 
## 
## Coefficients: (14 not defined because of singularities)
##                                                                                                     Estimate
## (Intercept)                                                                                          4.53122
## MFQSelf_c                                                                                            2.23461
## `MFQSelf_c:inoutgroup_effects`                                                                            NA
## `MFQSelf_c:ident_c`                                                                                       NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                    NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                  NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                             NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                        NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                          NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                              NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                           NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                      NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                 NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`       NA
##                                                                                                     Std. Error
## (Intercept)                                                                                            0.08175
## MFQSelf_c                                                                                              0.28187
## `MFQSelf_c:inoutgroup_effects`                                                                              NA
## `MFQSelf_c:ident_c`                                                                                         NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                      NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                    NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                               NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                          NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                       NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                  NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                             NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                        NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`         NA
##                                                                                                     t value
## (Intercept)                                                                                          55.428
## MFQSelf_c                                                                                             7.928
## `MFQSelf_c:inoutgroup_effects`                                                                           NA
## `MFQSelf_c:ident_c`                                                                                      NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                 NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                       NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                         NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                    NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                               NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                             NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                          NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`      NA
##                                                                                                                 Pr(>|t|)
## (Intercept)                                                                                         < 0.0000000000000002
## MFQSelf_c                                                                                             0.0000000000000094
## `MFQSelf_c:inoutgroup_effects`                                                                                        NA
## `MFQSelf_c:ident_c`                                                                                                   NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                              NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                                         NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                                    NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                                      NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                 NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                            NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                          NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                       NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                  NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                             NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                   NA
##                                                                                                        
## (Intercept)                                                                                         ***
## MFQSelf_c                                                                                           ***
## `MFQSelf_c:inoutgroup_effects`                                                                         
## `MFQSelf_c:ident_c`                                                                                    
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                 
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                               
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                          
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                     
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                       
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                  
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                             
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                           
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                        
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                   
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`              
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.84 on 667 degrees of freedom
##   (5 observations deleted due to missingness)
## Multiple R-squared:  0.08611,    Adjusted R-squared:  0.08474 
## F-statistic: 62.85 on 1 and 667 DF,  p-value: 0.000000000000009398

ggplot(downstream_coeffs,aes(MFQSelf_c, Treatment)) +
  geom_smooth(method = "lm") +
  labs(title = "Treatment predicted by projection coefficient",
       subtitle = "Coefficient from initial study controlling for stereotyping",
       x = "Projection coefficient") +
  theme_minimal()

Treatment with difference scores too

treat_codiff <- lm(Treatment ~ MFQSelf_c*proj_diff, coeffs_and_diffs)
summary(treat_codiff)

## 
## Call:
## lm(formula = Treatment ~ MFQSelf_c * proj_diff, data = coeffs_and_diffs)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -4.9667 -1.3728  0.4417  1.4476  3.9042 
## 
## Coefficients:
##                     Estimate Std. Error t value             Pr(>|t|)    
## (Intercept)           5.0318     0.1330  37.833 < 0.0000000000000002 ***
## MFQSelf_c             1.9712     0.4591   4.294          0.000020189 ***
## proj_diff            -0.7737     0.1503  -5.149          0.000000345 ***
## MFQSelf_c:proj_diff  -0.2071     0.5857  -0.354                0.724    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.797 on 665 degrees of freedom
##   (5 observations deleted due to missingness)
## Multiple R-squared:  0.1306, Adjusted R-squared:  0.1267 
## F-statistic:  33.3 on 3 and 665 DF,  p-value: < 0.00000000000000022

Both have main effects but the interaction is not significant.

Treatment with binding data

treat_coeff_bind <- lm(Treatment ~ MFQSelf_c*`MFQSelf_c:inoutgroup_effects`*`MFQSelf_c:ident_c`*`MFQSelf_c:inoutgroup_effects:ident_c`, data = downstream_coeffs_bind)
summary(treat_coeff_bind)

## 
## Call:
## lm(formula = Treatment ~ MFQSelf_c * `MFQSelf_c:inoutgroup_effects` * 
##     `MFQSelf_c:ident_c` * `MFQSelf_c:inoutgroup_effects:ident_c`, 
##     data = downstream_coeffs_bind)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -4.6927 -1.4613  0.4208  1.5608  3.1452 
## 
## Coefficients: (14 not defined because of singularities)
##                                                                                                     Estimate
## (Intercept)                                                                                          4.51005
## MFQSelf_c                                                                                            2.70685
## `MFQSelf_c:inoutgroup_effects`                                                                            NA
## `MFQSelf_c:ident_c`                                                                                       NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                    NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                  NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                             NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                        NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                          NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                              NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                           NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                      NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                 NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`       NA
##                                                                                                     Std. Error
## (Intercept)                                                                                            0.09391
## MFQSelf_c                                                                                              0.47303
## `MFQSelf_c:inoutgroup_effects`                                                                              NA
## `MFQSelf_c:ident_c`                                                                                         NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                      NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                    NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                               NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                          NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                       NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                  NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                             NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                        NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`         NA
##                                                                                                     t value
## (Intercept)                                                                                          48.026
## MFQSelf_c                                                                                             5.722
## `MFQSelf_c:inoutgroup_effects`                                                                           NA
## `MFQSelf_c:ident_c`                                                                                      NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                 NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                       NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                         NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                    NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                               NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                             NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                          NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`      NA
##                                                                                                                 Pr(>|t|)
## (Intercept)                                                                                         < 0.0000000000000002
## MFQSelf_c                                                                                                   0.0000000159
## `MFQSelf_c:inoutgroup_effects`                                                                                        NA
## `MFQSelf_c:ident_c`                                                                                                   NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                              NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                                         NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                                    NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                                      NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                 NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                            NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                          NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                       NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                  NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                             NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                   NA
##                                                                                                        
## (Intercept)                                                                                         ***
## MFQSelf_c                                                                                           ***
## `MFQSelf_c:inoutgroup_effects`                                                                         
## `MFQSelf_c:ident_c`                                                                                    
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                 
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                               
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                          
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                     
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                       
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                  
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                             
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                           
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                        
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                   
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`              
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.879 on 667 degrees of freedom
## Multiple R-squared:  0.0468, Adjusted R-squared:  0.04537 
## F-statistic: 32.75 on 1 and 667 DF,  p-value: 0.00000001587

ggplot(downstream_coeffs_bind,aes(MFQSelf_c, Treatment)) +
  geom_smooth(method = "lm") +
  labs(title = "Treatment predicted by projection coefficient",
       subtitle = "Coefficient from initial study with binding data only controlling for stereotyping",
       x = "Projection coefficient") +
  theme_minimal()

Treatment with individuating data

treat_coeff_indiv <- lm(Treatment ~ MFQSelf_c*`MFQSelf_c:inoutgroup_effects`*`MFQSelf_c:ident_c`*`MFQSelf_c:inoutgroup_effects:ident_c`, data = downstream_coeffs_indiv)
summary(treat_coeff_indiv)

## 
## Call:
## lm(formula = Treatment ~ MFQSelf_c * `MFQSelf_c:inoutgroup_effects` * 
##     `MFQSelf_c:ident_c` * `MFQSelf_c:inoutgroup_effects:ident_c`, 
##     data = downstream_coeffs_indiv)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -5.3567 -1.4758  0.4083  1.6451  3.2494 
## 
## Coefficients: (14 not defined because of singularities)
##                                                                                                     Estimate
## (Intercept)                                                                                           4.3638
## MFQSelf_c                                                                                             2.3044
## `MFQSelf_c:inoutgroup_effects`                                                                            NA
## `MFQSelf_c:ident_c`                                                                                       NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                    NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                  NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                             NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                        NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                          NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                              NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                           NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                      NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                 NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`       NA
##                                                                                                     Std. Error
## (Intercept)                                                                                             0.1140
## MFQSelf_c                                                                                               0.4158
## `MFQSelf_c:inoutgroup_effects`                                                                              NA
## `MFQSelf_c:ident_c`                                                                                         NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                      NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                    NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                               NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                          NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                       NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                  NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                             NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                        NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`         NA
##                                                                                                     t value
## (Intercept)                                                                                          38.265
## MFQSelf_c                                                                                             5.542
## `MFQSelf_c:inoutgroup_effects`                                                                           NA
## `MFQSelf_c:ident_c`                                                                                      NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                 NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                       NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                         NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                    NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                               NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                             NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                          NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`      NA
##                                                                                                                 Pr(>|t|)
## (Intercept)                                                                                         < 0.0000000000000002
## MFQSelf_c                                                                                                   0.0000000431
## `MFQSelf_c:inoutgroup_effects`                                                                                        NA
## `MFQSelf_c:ident_c`                                                                                                   NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                              NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                                         NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                                    NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                                      NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                 NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                            NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                          NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                       NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                  NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                             NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                   NA
##                                                                                                        
## (Intercept)                                                                                         ***
## MFQSelf_c                                                                                           ***
## `MFQSelf_c:inoutgroup_effects`                                                                         
## `MFQSelf_c:ident_c`                                                                                    
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                 
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                               
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                          
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                     
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                       
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                  
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                             
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                           
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                        
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                   
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`              
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.882 on 667 degrees of freedom
## Multiple R-squared:  0.04402,    Adjusted R-squared:  0.04259 
## F-statistic: 30.71 on 1 and 667 DF,  p-value: 0.00000004311

ggplot(downstream_coeffs_indiv,aes(MFQSelf_c, Treatment)) +
  geom_smooth(method = "lm") +
  labs(title = "Treatment predicted by projection coefficient",
       subtitle = "Coefficient from initial study with individuating data only controlling for stereotyping",
       x = "Projection coefficient") +
  theme_minimal()

Treatment effects with individuating and binding data look pretty much the same as the overall model.

Validity predicted by coefficients

valid_coeffs <- lm(Valid ~ MFQSelf_c*`MFQSelf_c:inoutgroup_effects`*`MFQSelf_c:ident_c`*`MFQSelf_c:inoutgroup_effects:ident_c`, data = downstream_coeffs)
summary(valid_coeffs)

## 
## Call:
## lm(formula = Valid ~ MFQSelf_c * `MFQSelf_c:inoutgroup_effects` * 
##     `MFQSelf_c:ident_c` * `MFQSelf_c:inoutgroup_effects:ident_c`, 
##     data = downstream_coeffs)
## 
## Residuals:
##    Min     1Q Median     3Q    Max 
## -4.839 -1.548  0.519  1.595  3.420 
## 
## Coefficients: (14 not defined because of singularities)
##                                                                                                     Estimate
## (Intercept)                                                                                          4.53539
## MFQSelf_c                                                                                            2.14270
## `MFQSelf_c:inoutgroup_effects`                                                                            NA
## `MFQSelf_c:ident_c`                                                                                       NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                    NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                  NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                             NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                        NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                          NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                              NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                           NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                      NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                 NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`       NA
##                                                                                                     Std. Error
## (Intercept)                                                                                            0.08455
## MFQSelf_c                                                                                              0.29152
## `MFQSelf_c:inoutgroup_effects`                                                                              NA
## `MFQSelf_c:ident_c`                                                                                         NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                      NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                    NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                               NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                          NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                       NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                  NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                             NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                        NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`         NA
##                                                                                                     t value
## (Intercept)                                                                                           53.64
## MFQSelf_c                                                                                              7.35
## `MFQSelf_c:inoutgroup_effects`                                                                           NA
## `MFQSelf_c:ident_c`                                                                                      NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                 NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                       NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                         NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                    NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                               NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                             NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                          NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`      NA
##                                                                                                                 Pr(>|t|)
## (Intercept)                                                                                         < 0.0000000000000002
## MFQSelf_c                                                                                              0.000000000000582
## `MFQSelf_c:inoutgroup_effects`                                                                                        NA
## `MFQSelf_c:ident_c`                                                                                                   NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                              NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                                         NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                                    NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                                      NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                 NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                            NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                          NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                       NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                  NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                             NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                   NA
##                                                                                                        
## (Intercept)                                                                                         ***
## MFQSelf_c                                                                                           ***
## `MFQSelf_c:inoutgroup_effects`                                                                         
## `MFQSelf_c:ident_c`                                                                                    
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                 
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                               
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                          
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                     
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                       
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                  
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                             
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                           
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                        
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                   
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`              
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.903 on 667 degrees of freedom
##   (5 observations deleted due to missingness)
## Multiple R-squared:  0.07493,    Adjusted R-squared:  0.07354 
## F-statistic: 54.02 on 1 and 667 DF,  p-value: 0.0000000000005815

ggplot(downstream_coeffs,aes(MFQSelf_c, Valid)) +
  geom_smooth(method = "lm") +
  labs(title = "Validity of target's position predicted by projection coefficient",
       subtitle = "Coefficient from initial study controlling for stereotyping",
       x = "Projection coefficient",
       y = "Validity") +
  theme_minimal()

Validity with difference scores too

valid_codiff <- lm(Valid ~ MFQSelf_c*proj_diff, coeffs_and_diffs)
summary(valid_codiff)

## 
## Call:
## lm(formula = Valid ~ MFQSelf_c * proj_diff, data = coeffs_and_diffs)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -4.9623 -1.3315  0.5359  1.4941  4.0047 
## 
## Coefficients:
##                     Estimate Std. Error t value             Pr(>|t|)    
## (Intercept)           5.0766     0.1377  36.863 < 0.0000000000000002 ***
## MFQSelf_c             1.6560     0.4754   3.484             0.000527 ***
## proj_diff            -0.8204     0.1556  -5.273          0.000000181 ***
## MFQSelf_c:proj_diff   0.1479     0.6065   0.244             0.807380    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.861 on 665 degrees of freedom
##   (5 observations deleted due to missingness)
## Multiple R-squared:  0.1179, Adjusted R-squared:  0.1139 
## F-statistic: 29.63 on 3 and 665 DF,  p-value: < 0.00000000000000022

Both have main effects but the interaction is not significant.

Validity with binding data

valid_coeff_bind <- lm(Valid ~ MFQSelf_c*`MFQSelf_c:inoutgroup_effects`*`MFQSelf_c:ident_c`*`MFQSelf_c:inoutgroup_effects:ident_c`, data = downstream_coeffs_bind)
summary(valid_coeff_bind)

## 
## Call:
## lm(formula = Valid ~ MFQSelf_c * `MFQSelf_c:inoutgroup_effects` * 
##     `MFQSelf_c:ident_c` * `MFQSelf_c:inoutgroup_effects:ident_c`, 
##     data = downstream_coeffs_bind)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -4.7196 -1.4663  0.4361  1.6767  2.9226 
## 
## Coefficients: (14 not defined because of singularities)
##                                                                                                     Estimate
## (Intercept)                                                                                          4.54937
## MFQSelf_c                                                                                            2.32294
## `MFQSelf_c:inoutgroup_effects`                                                                            NA
## `MFQSelf_c:ident_c`                                                                                       NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                    NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                  NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                             NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                        NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                          NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                              NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                           NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                      NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                 NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`       NA
##                                                                                                     Std. Error
## (Intercept)                                                                                            0.09725
## MFQSelf_c                                                                                              0.48986
## `MFQSelf_c:inoutgroup_effects`                                                                              NA
## `MFQSelf_c:ident_c`                                                                                         NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                      NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                    NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                               NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                          NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                       NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                  NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                             NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                        NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`         NA
##                                                                                                     t value
## (Intercept)                                                                                          46.780
## MFQSelf_c                                                                                             4.742
## `MFQSelf_c:inoutgroup_effects`                                                                           NA
## `MFQSelf_c:ident_c`                                                                                      NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                 NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                       NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                         NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                    NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                               NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                             NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                          NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`      NA
##                                                                                                                 Pr(>|t|)
## (Intercept)                                                                                         < 0.0000000000000002
## MFQSelf_c                                                                                                     0.00000259
## `MFQSelf_c:inoutgroup_effects`                                                                                        NA
## `MFQSelf_c:ident_c`                                                                                                   NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                              NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                                         NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                                    NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                                      NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                 NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                            NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                          NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                       NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                  NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                             NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                   NA
##                                                                                                        
## (Intercept)                                                                                         ***
## MFQSelf_c                                                                                           ***
## `MFQSelf_c:inoutgroup_effects`                                                                         
## `MFQSelf_c:ident_c`                                                                                    
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                 
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                               
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                          
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                     
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                       
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                  
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                             
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                           
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                        
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                   
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`              
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.946 on 667 degrees of freedom
## Multiple R-squared:  0.03261,    Adjusted R-squared:  0.03116 
## F-statistic: 22.49 on 1 and 667 DF,  p-value: 0.000002589

ggplot(downstream_coeffs_bind,aes(MFQSelf_c, Valid)) +
  geom_smooth(method = "lm") +
  labs(title = "Validity predicted by projection coefficient",
       subtitle = "Coefficient from initial study with binding data only controlling for stereotyping",
       x = "Projection coefficient",
       y = "Validity") +
  theme_minimal()

Validity with individuating data

valid_coeff_indiv <- lm(Valid ~ MFQSelf_c*`MFQSelf_c:inoutgroup_effects`*`MFQSelf_c:ident_c`*`MFQSelf_c:inoutgroup_effects:ident_c`, data = downstream_coeffs_indiv)
summary(valid_coeff_indiv)

## 
## Call:
## lm(formula = Valid ~ MFQSelf_c * `MFQSelf_c:inoutgroup_effects` * 
##     `MFQSelf_c:ident_c` * `MFQSelf_c:inoutgroup_effects:ident_c`, 
##     data = downstream_coeffs_indiv)
## 
## Residuals:
##    Min     1Q Median     3Q    Max 
## -4.622 -1.452  0.492  1.731  3.061 
## 
## Coefficients: (14 not defined because of singularities)
##                                                                                                     Estimate
## (Intercept)                                                                                           4.4421
## MFQSelf_c                                                                                             1.8915
## `MFQSelf_c:inoutgroup_effects`                                                                            NA
## `MFQSelf_c:ident_c`                                                                                       NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                    NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                  NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                             NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                        NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                          NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                              NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                           NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                      NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                 NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`       NA
##                                                                                                     Std. Error
## (Intercept)                                                                                             0.1182
## MFQSelf_c                                                                                               0.4310
## `MFQSelf_c:inoutgroup_effects`                                                                              NA
## `MFQSelf_c:ident_c`                                                                                         NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                      NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                    NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                               NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                          NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                       NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                  NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                             NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                        NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`         NA
##                                                                                                     t value
## (Intercept)                                                                                          37.579
## MFQSelf_c                                                                                             4.389
## `MFQSelf_c:inoutgroup_effects`                                                                           NA
## `MFQSelf_c:ident_c`                                                                                      NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                   NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                 NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                            NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                       NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                         NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                    NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                               NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                             NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                          NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                     NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`      NA
##                                                                                                                 Pr(>|t|)
## (Intercept)                                                                                         < 0.0000000000000002
## MFQSelf_c                                                                                                      0.0000133
## `MFQSelf_c:inoutgroup_effects`                                                                                        NA
## `MFQSelf_c:ident_c`                                                                                                   NA
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                                NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                                              NA
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                                         NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                                    NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                                      NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                 NA
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                            NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                          NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                       NA
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                                  NA
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                             NA
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                   NA
##                                                                                                        
## (Intercept)                                                                                         ***
## MFQSelf_c                                                                                           ***
## `MFQSelf_c:inoutgroup_effects`                                                                         
## `MFQSelf_c:ident_c`                                                                                    
## `MFQSelf_c:inoutgroup_effects:ident_c`                                                                 
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`                                                               
## MFQSelf_c:`MFQSelf_c:ident_c`                                                                          
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                                     
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects:ident_c`                                                       
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                                  
## `MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                             
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`                                           
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:inoutgroup_effects:ident_c`                        
## MFQSelf_c:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`                                   
## `MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`              
## MFQSelf_c:`MFQSelf_c:inoutgroup_effects`:`MFQSelf_c:ident_c`:`MFQSelf_c:inoutgroup_effects:ident_c`    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.951 on 667 degrees of freedom
## Multiple R-squared:  0.02807,    Adjusted R-squared:  0.02661 
## F-statistic: 19.26 on 1 and 667 DF,  p-value: 0.00001326

ggplot(downstream_coeffs_indiv,aes(MFQSelf_c, Valid)) +
  geom_smooth(method = "lm") +
  labs(title = "Validity predicted by projection coefficient",
       subtitle = "Coefficient from initial study with individuating data only controlling for stereotyping",
       x = "Projection coefficient",
       y = "Validity") +
  theme_minimal()

Models using the individuating and binding data look the same as the overall model.

Look at how they respond for in-group to address demand?

Does this mean looking at whether they are approve or disapprove of the condition in F1 and F2? Could do liberal leaning vs conservative leaning, minus PB and third party

Alphas for the MFQ

In the original output for initial study and follow-up 1, already done.